Quantum-Inspired Genetic Optimization for Patient Scheduling in Radiation Oncology

Akira SaiToh; Amit Sawant; Arezoo Modiri; Robabeh Rahimi

REVIEW 4 major objections 4 minor 26 references

Quantum-Inspired Genetic Optimization for Patient Scheduling in Radiation Oncology

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A quantum-inspired genetic algorithm can produce proton-therapy schedules as good as a classical one while needing roughly a third to half the chromosomes, at the cost of longer emulation time.

desk verdict The claimed population-size advantage is an artifact of counting chromosomes instead of stored amplitudes, and the paper's own conclusion is more modest than its abstract. read the letter →

arxiv 2506.04328 v1 pith:V5YL6VMI submitted 2025-06-04 cs.NE physics.med-ph

Akira SaiToh , Arezoo Modiri , Amit Sawant , Robabeh Rahimi This is my paper

classification cs.NEphysics.med-ph

keywords quantum-inspiredgeneticalgorithmpatientschedulingprotontherapyradiationoncologysuperpositionchromosomeamplitudeamplificationpopulationsizeevolutionaryoptimization

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

A quantum-inspired genetic algorithm can produce daily proton-therapy schedules as good as a classical genetic algorithm while using far fewer chromosomes, according to numerical experiments on a 12-patient and a 72-patient instance. The quantum version caps its population at 50 or 70 chromosomes, while the classical version needs 150 or 250, yet both reach comparable or better fitness. The authors are explicit that this population-size advantage does not translate into a runtime advantage in classical emulation: the quantum-inspired runs take 2.4 to 5.7 times longer. If the resource advantage survives on genuine quantum hardware, the approach could make large clinic scheduling problems tractable with far fewer stored candidate schedules.

What carries the argument

The central device is the quantum chromosome: an array with one cell per gantry and time slot in which each cell stores a superposition over available patient IDs and a superposition over the eight gantry statuses, rather than a single value. The repair operation acts as a simulated high-dimensional unitary rotation: it boosts amplitudes of desirable statuses and IDs by a factor of ten (down to a $\sqrt{1/4}$ floor) and renormalizes, steering a small population toward clinically feasible schedules. This superposition representation is what lets a small population explore many schedule variants per cell, and it is also what makes classical emulation slow.

What would settle it

Run the same parameter sweeps with a fixed memory budget instead of a fixed chromosome cap: allow each algorithm the same total number of stored values, counting every amplitude $a_j$ and $b_k$ in every cell of every quantum chromosome. If a classical genetic algorithm then matches or exceeds the quantum-inspired fitness on both instances, the population-size advantage is an artifact of the chosen fairness metric. A simpler check is to record best fitness versus total allocated amplitudes per generation for both algorithms.

Watch

Extended reading notes

Core claim

The paper proposes a quantum-inspired genetic algorithm for daily proton-therapy scheduling in which each cell of a chromosome, for a given gantry $\square$ and time slot $t$, holds a normalized superposition $\sum_j a_j(\square,t)|j\rangle$ over patient IDs and a normalized superposition $\sum_k b_k(\square,t)|k\rangle$ over gantry statuses, instead of a single ID and status. Fitness is evaluated by simulated non-demolition projective measurements of these superpositions, and the repair operation amplifies the amplitudes of clinically desirable IDs and statuses by a factor of ten, down to a floor of $\sqrt{1/4}$, before renormalizing. The paper's numerical claim is that this representation reaches clinically feasible schedules with population caps of 50 versus 150 on a medium-size instance and 70 versus 250 on a large-size practical instance, with comparable or higher fitness in parameter sweeps (for example, 4836 versus 3913 on the large instance) while taking longer because classical emulation must manipulate many amplitudes.

Load-bearing premise

The claimed advantage treats the number of chromosomes as the resource being measured; if instead one counts stored amplitudes per cell or wall-clock time, the 'clear advantage in population size' may shrink or disappear.

Editorial extensions

If this is right

If the population-size advantage holds, a quantum-inspired scheduler could match classical schedule quality with roughly one-third to one-half the chromosomes, easing the memory demands of storing full daily schedules.
The same chromosome design transfers to photon-therapy scheduling by changing the gantry status durations in the model, so the approach is not specific to proton therapy.
On a real quantum computer, the classical emulation overhead would be removed, although the paper estimates the required qubit count is far beyond near-term hardware.
The classical genetic algorithm remains a stable fallback for clinics without high-performance computing, since it steadily produces high-fitness schedules.
The paper's omission of group-migration and pair-swap strategies suggests those extras are not necessary for this scheduling problem, simplifying future quantum-inspired implementations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A fairness issue is left implicit: the comparison counts chromosomes, but each quantum chromosome stores many amplitudes per cell; if population size is replaced by total stored values or wall-clock time as the resource metric, the reported advantage may shrink or reverse. A testable extension would rerun the sweeps under a fixed memory budget.
The repair operation is a strong, problem-specific local search that boosts desired amplitudes by a factor of ten; its contribution to the fitness gain is not separated from the contribution of the superposition representation itself, so part of the observed advantage may come from repair rather than quantum-inspired encoding.
On genuine quantum hardware, the simulated non-demolition measurements would become real measurements, and the amplitude amplification would need to be implemented as coherent unitary operations; whether the smaller population bound yields an end-to-end speedup depends on the overhead of those operations, which the paper does not claim to resolve.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The one thing to know: the abstract's “clear advantage in population size” is not established. The comparison counts chromosomes, but each quantum chromosome stores a superposition over patient IDs and gantry statuses per cell—np+ns amplitudes, versus two classical values. With np=72 and ns=8, that's 80 amplitudes per cell against 2, so the QI population holds roughly an order of magnitude more numbers per generation. Reported runtimes (1478 vs 261 s) match that overhead. And no experiment sweeps Nmax for either algorithm, so we can't tell whether the classical GA would also do fine at 50 or 70 chromosomes. The headline claim is a metric artifact, not a demonstrated resource advantage.

Now the credit. The paper does something genuinely new: tailoring a quantum-inspired genetic algorithm to proton therapy scheduling with a chromosome that superposes patient IDs and gantry statuses. The 72-patient instance is a real-world-scale test, and the authors are honest about the long runtimes and the classical GA's stability. They also candidly state in the conclusion that the quantum-inspired algorithm's merit is currently insignificant. The fitness model and the classical baseline are credible; there's no circularity or fitted-to-data derivation.

The soft spots are concentrated in the empirical comparison. Best-of-sweep points are compared without repeated seeds or error bars. One data set (rc=0.23) is excluded post hoc as an outlier. The parameter sweeps use different effective budgets for the two algorithms because of the amplitude overhead. These issues weaken the central claim but don't sink the application itself. With a fair resource metric and repeated runs, the conclusion might shift to “comparable fitness at comparable real cost” — which would still be a mild but legitimate result.

Who is this for? Researchers interested in quantum-inspired evolutionary algorithms applied to scheduling. It's a useful case study in how easy it is to miscompare resource usage when quantum states are emulated classically.

Recommendation: send it to peer review, but the referee should demand an Nmax sweep for both algorithms, a resource metric that counts stored amplitudes or memory/CPU cost, and repeated seeds with variance reporting. The current version's framing needs major revision, but the underlying application is legitimate and worth referee time, not a desk reject.

Referee Report

4 major / 4 minor

Summary. The manuscript proposes a classical genetic algorithm (CGA) and a quantum-inspired genetic algorithm (QIGA) for daily proton-therapy patient scheduling. Each chromosome encodes patient IDs and gantry statuses over time slots for multiple gantries; the QIGA version represents each cell as a superposition over patient IDs and gantry statuses via Eqs. (1)-(2), selects through simulated projective measurements, and uses amplitude-enhancing repair. Numerical experiments on a 12-patient medium case and a 72-patient large case compare single runs at Nmax=50 vs 150 and Nmax=70 vs 250, plus parameter sweeps over evolution rates. The authors claim a 'clear advantage in population size' for QIGA while reporting longer runtimes, and conclude in Section 6 that its merit is 'insignificant in comparison to the classical counterpart currently.'

Significance. If the population-size advantage were established under a fair resource metric, this would be a useful contribution to quantum-inspired evolutionary computation for scheduling. The quantum chromosome design is a nontrivial constructive extension of qubit-based QIGA to a superposition representation for a practical operations-research problem, and the manuscript is transparent about runtimes and about the stability of the classical GA. The paper contains no fitted derivation and no circularity; the QI components are constructive heuristics. However, the central claim is currently unsupported because chromosome count is used as the resource metric despite each QI chromosome storing np+ns amplitudes per cell, and because the comparison lacks Nmax sweeps, repeated seeds, and error bars. The practical significance is therefore not yet demonstrated.

major comments (4)

[§4.1-4.2, Tables 3 and 7, Figs. 6-10] The claimed 'clear advantage in population size' is not supported by the reported experiments. The medium and large comparisons each use a single QI run at one cap (Nmax=50 or 70) and a single classical run at another cap (Nmax=150 or 250), with no sweep of Nmax for either algorithm; the fitness values at these caps are comparable or favor the classical GA (674 vs 728 medium; 4499 vs 3818 large), and the classical GA converges faster in the large case. Without varying Nmax, the data cannot show that QI requires fewer chromosomes or that the classical GA would not reach the same fitness at Nmax=50 or 70.
[§3.2.1, Eqs. (1)-(2), Tables 4 and 8] Chromosome count is not a fair resource metric for this comparison. A classical cell stores one patient ID and one gantry status, whereas each QI cell stores np+ns complex amplitudes for the superpositions in Eqs. (1)-(2); with np=12/72 and ns=8 this is 20 or 80 stored numbers per cell versus 2. Under the caps used (50 vs 150; 70 vs 250), the QI population stores roughly 3 and 11 times as many numbers per generation as the classical population. The reported runtimes (19.7 vs 8.29 s; 1478 vs 261 s) are consistent with this overhead. Thus the abstract's 'small resource requirement' and 'clear advantage in population size' are artifacts of the chosen metric rather than demonstrated resource advantages.
[§4.1.1 and §4.2.1, Tables 5, 6, 9, 10] The parameter-sweep comparisons are not strong evidence for QI advantage. Each sweep point appears to be a single run with no repeated seeds or error bars; the 'outlier' exclusion of rc=0.23 in §4.1.1 is post hoc and has no stated statistical criterion; and the QI fitness distributions are extremely unstable (e.g., large-case min -27106, standard deviation 6.73e3, with negative average fitness in Table 9). Comparing only the best fitness values of the sweeps (893 vs 915 medium; 4836 vs 3913 large) therefore overstates robustness and does not establish a consistent advantage.
[Abstract and Section 6] The abstract's claims of 'clear advantage in population size' and 'small resource requirement' are inconsistent with the paper's own conclusion in Section 6 that 'its merit was insignificant in comparison to the classical counterpart currently' and with the reported runtime overhead in Section 5. The numerical results support at most a qualified, metric-specific observation, and the manuscript should be revised to present the findings with this qualification.

minor comments (4)

[§3.2.1, Eqs. (1)-(2)] In the version under review, the mathematics in Eqs. (1)-(2) contains placeholder glyphs, which makes it difficult to verify the normalization conditions and the dimension of the superpositions; please ensure all symbols render correctly.
[§3.2.5] The phrase 'bypassed simulation of a high-dimensional unitary rotation' should be clarified: the amplitude enhancement by a factor of ten is a heuristic weight update, not an actual unitary operation, and the no-cloning issue is avoided only because the emulation is classical.
[§4.2, Table 7] The initial population sizes differ (40 for the classical algorithm, 10 for the quantum-inspired one); this asymmetry should be justified or controlled, since it may affect early convergence and the subsequent population-size comparison.
[Table 2] The penalty and benefit scores in Table 2 are presented without clinical justification; this is acceptable for an algorithmic comparison under a fixed model, but it limits the clinical interpretation of the absolute fitness values.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the comparison is an empirical benchmark, and the population-size claim is a resource-metric concern rather than a circular derivation.

full rationale

The paper's contributions are constructive GA heuristics plus numerical comparisons; it does not derive any quantity from fitted parameters and then 'predict' that same quantity. The fitness model in Table 2 is a fixed scoring objective used equally by both the classical and quantum-inspired algorithms, so there is no self-definitional loop in the optimization target. The 'clear advantage in population size' rests on comparing preset caps (QI Nmax=50 vs classical Nmax=150 for the medium case; QI Nmax=70 vs classical Nmax=250 for the large case), which may be an unfair resource metric because each quantum chromosome stores np+ns amplitudes per cell, but this is a benchmarking/validity weakness, not a circular reduction. The manuscript itself concedes in Sec. 6 that 'its merit was insignificant in comparison to the classical counterpart currently,' which undercuts the abstract's resource-superiority rhetoric but again is a correctness caveat, not circularity. The only author self-citation, ref. [26], appears in a forward-looking sentence about possible future use of real quantum computers ('A technique to use a real quantum computer for evolutionary computing [24-26] will then be useful') and is not load-bearing for any result in the paper. No uniqueness theorem is imported from the authors' prior work, no ansatz is smuggled in via self-citation, and no known result is merely renamed. The derivation chain is therefore self-contained: each algorithmic step is explicitly constructed and numerically tested rather than being equivalent to its own input.

Assumptions & free parameters 5 free parameters · 5 assumptions · 1 invented entities

The comparison is empirical, so the ledger is dominated by modeling choices rather than mathematical axioms. The main free parameters are the hand-set fitness scores, population caps, and evolution rates, with best values selected from sweeps. The domain assumptions define the scheduling model and the validity of the fitness function as a proxy for clinical feasibility; the ad hoc assumptions concern the quantum-inspired amplitude update and crossover ignoring no-cloning.

free parameters (5)

Population size caps (Nmax) = 50 (QI) and 150 (classical) for medium; 70 (QI) and 250 (classical) for large
The central population-size advantage is defined by these chosen caps; no sensitivity analysis shows QI works with these caps beyond the reported runs.
Penalty and benefit scores in Table 2 = Patient conflict 20, non-consecutive statuses 20, multiple treatments 28, interruption 12, time penalty 1.5, benefits…
Hand-set fitness coefficients define what counts as a good schedule; no calibration or justification from clinical outcomes is given.
Evolution rates (rs, rc, rm, rr) = Medium: 0.83, 0.27, 0.37, 0.85; large: 0.83, 0.37, 0.37, 0.85; variants from sweeps
Selected by parameter sweeps; best results are quoted after sweeping, so values are fit to the benchmark.
Repair amplitude boost factor = Factor 10 or until 1/4
Ad hoc choice for amplitude amplification; no derivation that it matches a unitary rotation in a real quantum setting.
Outlier exclusion threshold = rc=0.23 excluded
Post hoc exclusion criterion is not specified beyond 'obvious' separation; affects Table 5 statistics.

assumptions (5)

domain assumption Daily scheduling model: one session per patient, three gantries, fixed status transition sequence with durations in Table 1.
Defines the problem instance; BID is excluded, no offline imaging or shuttle conveyance is modeled, and durations are taken from Sakae et al. [7].
domain assumption The fitness function in Table 2 is a valid proxy for clinical schedule feasibility.
All quality comparisons rely on these hand-set scores; no validation against actual clinical operations or outcomes is provided.
domain assumption Simulated projection measurement with uniform random numbers is a faithful emulation of quantum measurement.
Used in Section 3.2.2 to convert superpositions to concrete IDs and statuses for fitness evaluation.
ad hoc to paper Amplitude enhancement by factor ten is equivalent to a high-dimensional unitary rotation in QIGA.
Section 3.2.5 calls it a bypassed simulation; no proof is given that it preserves the intended quantum-inspired dynamics.
ad hoc to paper Crossover can ignore the no-cloning theorem in this emulation.
Section 3.2.3 applies the classical crossover to quantum chromosomes; a real quantum implementation would need to address no-cloning.

invented entities (1)

Quantum chromosome storing superposed patient IDs and gantry statuses
purpose: Represent candidate schedules so amplitude amplification can bias cells toward high-fitness options.
Introduced in Section 3.2.1; only realized as classical arrays of amplitudes, not as physical quantum states, and no external falsifiable handle is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantum-Inspired Genetic Optimization for Patient Scheduling in Radiation Oncology." pith.science (2026). https://pith.science/paper/V5YL6VMI

@misc{pith2026250604328,
  author       = {Pith},
  title        = {Pith review of: Quantum-Inspired Genetic Optimization for Patient Scheduling in Radiation Oncology},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V5YL6VMI}},
  note         = {Machine review of arXiv:2506.04328}
}

read the original abstract

Among the genetic algorithms generally used for optimization problems in the recent decades, quantum-inspired variants are known for fast and high-fitness convergence and small resource requirement. Here the application to the patient scheduling problem in proton therapy is reported. Quantum chromosomes are tailored to possess the superposed data of patient IDs and gantry statuses. Selection and repair strategies are also elaborated for reliable convergence to a clinically feasible schedule although the employed model is not complex. Clear advantage in population size is shown over the classical counterpart in our numerical results for both a medium-size test case and a large-size practical problem instance. It is, however, observed that program run time is rather long for the large-size practical case, which is due to the limitation of classical emulation and demands the forthcoming true quantum computation. Our results also revalidate the stability of the conventional classical genetic algorithm.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

26 extracted references · 26 canonical work pages

[1]

Cancer and Radiation Therapy: Current Advances and Future Directions

Baskar R, Lee KA, Yeo R, Yeoh KW. Cancer and Radiation Therapy: Current Advances and Future Directions. Int J Med Sci. 2012; 9(3):193-199

work page 2012
[2]

A Critical Review of Radiation Therapy: From Particle Beam Therapy (Proton, Carbon, and BNCT) to Beyond

Matsumoto Y, Fukumitsu N, Ishikawa H, Nakai K, Sakurai H. A Critical Review of Radiation Therapy: From Particle Beam Therapy (Proton, Carbon, and BNCT) to Beyond. J Pers Med. 2021; 11:825

work page 2021
[3]

Review of proton therapy – Current status and future directions

Mohan RA. Review of proton therapy – Current status and future directions. Prec Radiat Oncol. 2022; 6:164–176

work page 2022
[4]

Process Modeling a Radiation Oncology Clinic Workflow From Therapeutic Simulation to Treatment: Identifying Impending Strain and Possible Treatment Delays

Chowdhry VK, Simpson NC. Process Modeling a Radiation Oncology Clinic Workflow From Therapeutic Simulation to Treatment: Identifying Impending Strain and Possible Treatment Delays. Adv Radiat Oncol. 2023; 8(6):101261

work page 2023
[5]

Complexity and Approximation - Combinatorial Optimization Problems and Their Approximability Properties 2nd Ed

Ausiello G, Crescenzi P, Gambosi G, Kann V, Marchetti-Spaccamela A, Protasi M. Complexity and Approximation - Combinatorial Optimization Problems and Their Approximability Properties 2nd Ed. Berlin: Springer; 2003

work page 2003
[6]

In-gantry or remote patient positioning? Monte Carlo simulations for proton therapy centers of different sizes

Fava G, et al. In-gantry or remote patient positioning? Monte Carlo simulations for proton therapy centers of different sizes. Radiotherapy and Oncology. 2012; 103:18-24

work page 2012
[7]

Modeling of daily operation in proton radiotherapy by Monte Carlo method

Sakae T, Tsunashima Y, Terunuma T, Sato M. Modeling of daily operation in proton radiotherapy by Monte Carlo method. Jpn J Med Phys. 2003; 23(2):147-156. Japanese

work page 2003
[8]

Healthcare scheduling in 15 optimization context: a review

Abdalkareem ZA, Amir A, Al-Betar MA, Ekhan P, Hammouri AI. Healthcare scheduling in 15 optimization context: a review. Health Technol. 2021; 11:445-469

work page 2021

Show all 26 references

[9]

Genetic Algorithms in Search, Optimization, and Machine Learning

Goldberg DE. Genetic Algorithms in Search, Optimization, and Machine Learning. Massachusetts: Addison Wesley; 1989

1989
[10]

A Diversity Study in Genetic Algorithms for Job Shop Scheduling Problems

Brizuela CA, Sannomiya N. A Diversity Study in Genetic Algorithms for Job Shop Scheduling Problems. Proceedings of GECCO99; 1999. pp. 75-82

1999
[11]

Evolutionary Scheduling: A Review

Hart E, Ross P, Corne D. Evolutionary Scheduling: A Review. Genetic Program Evol Mach. 2005; 6:191-220

2005
[12]

Quantum-inspired genetic algorithms

Narayanan A, Moore M. Quantum-inspired genetic algorithms. Proceedings of IEEE Int. Conf. Evol. Comput.: 1996. pp. 61-66

1996
[13]

Quantum-Inspired Evolutionary Algorithm for a Class of Combinatorial Optimization

Han K-H, Kim J-H. Quantum-Inspired Evolutionary Algorithm for a Class of Combinatorial Optimization. IEEE Trans. Evol. Comput. 2002; 6(6):580-593

2002
[14]

Pair Swap Strategy in Quantum-Inspired Evolutionary Algorithm

Nakayama S, Imabeppu T, Ono S. Pair Swap Strategy in Quantum-Inspired Evolutionary Algorithm. Proceedings of GECCO2006: 2006. in late-breaking papers session. See also ibid. Study on Quantum-Inspired Evolutionary Algorithm. IPSJ SIG Tech. Rep. 2007; 2007-AL-111 (8):57-64. Japanese

2006
[15]

Quantum-Inspired Genetic Algorithm for Resource-Constrained Project-Scheduling

Saad HMH, Chakrabortty RK, Elsayed S, Ryan MJ. Quantum-Inspired Genetic Algorithm for Resource-Constrained Project-Scheduling. IEEE Access 2021; 9:38488-38502

2021
[16]

MRI-LINAC: A transformative technology in radiation oncology

Ng J et al. MRI-LINAC: A transformative technology in radiation oncology. Front Oncol 2023; 13:1117874

2023
[17]

Introducing a new treatment flow using the Shuttle Treatment Table

B dot Medical Inc. Introducing a new treatment flow using the Shuttle Treatment Table. 37th JASTRO Meeting; 2024 Nov 21-23; Yokohama, Japan: seminar in Luncheon Seminar 2. See also https://bdotmed.co.jp/en/product/throup/

2024
[18]

A Heuristic Genetic Algorithm for Solving Resource Allocation Problems

Lee Z-J, et al. A Heuristic Genetic Algorithm for Solving Resource Allocation Problems. Knowl. Inf. Sys. 2003; 5:503-11

2003
[19]

Improved representation and genetic operators for linear genetic programming for automated program repair

Oliveira VPL, et al. Improved representation and genetic operators for linear genetic programming for automated program repair. Empir. Soft. Eng. 2018; 23:2980-3006

2018
[20]

Quantum vs classical genetic algorithms: A numerical comparison shows faster convergence

Ibarrondo R, Gatti G, Sanz M. Quantum vs classical genetic algorithms: A numerical comparison shows faster convergence. Proceedings of 2022 IEEE Symposium Series on Computational Intelligence (SSCI 2022): 2022. pp. 947-954

2022
[21]

Quantum Genetic Algorithm for Highly Constrained Optimization Problems

Sabaawi AMA, et al. Quantum Genetic Algorithm for Highly Constrained Optimization Problems. Infocom. J. 2023; 15(3):63-71

2023
[22]

Characterizing Quantum Supremacy in Near-Term Devices

Boixo S, et al. Characterizing Quantum Supremacy in Near-Term Devices. Nat. Phys. 2018; 14:595-600

2018
[23]

Quantum Computing in the NISQ era and beyond

Preskill J. Quantum Computing in the NISQ era and beyond. Quantum 2018; 2:79

2018
[24]

Implementing quantum genetic algorithms: a solution based on Grover’s algorithm

Udrescu M, Prodan L, Vlăduţiu M. Implementing quantum genetic algorithms: a solution based on Grover’s algorithm. Proceedings of the 3rd Conference on Computing Frontiers: 2006. pp. 71–81

2006
[25]

Can quantum search accelerate evolutionary algorithms? Proceedings of GECCO2010: 2010

Johannsen D, Kuru PP, Lengler J. Can quantum search accelerate evolutionary algorithms? Proceedings of GECCO2010: 2010. pp. 1433–1440

2010
[26]

A quantum genetic algorithm with quantum crossover and mutation operations

SaiToh A, Rahimi R, Nakahara M. A quantum genetic algorithm with quantum crossover and mutation operations. Quantum Inf. Process. 2014; 13:737-55

2014

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Cancer and Radiation Therapy: Current Advances and Future Directions

Baskar R, Lee KA, Yeo R, Yeoh KW. Cancer and Radiation Therapy: Current Advances and Future Directions. Int J Med Sci. 2012; 9(3):193-199

work page 2012

[2] [2]

A Critical Review of Radiation Therapy: From Particle Beam Therapy (Proton, Carbon, and BNCT) to Beyond

Matsumoto Y, Fukumitsu N, Ishikawa H, Nakai K, Sakurai H. A Critical Review of Radiation Therapy: From Particle Beam Therapy (Proton, Carbon, and BNCT) to Beyond. J Pers Med. 2021; 11:825

work page 2021

[3] [3]

Review of proton therapy – Current status and future directions

Mohan RA. Review of proton therapy – Current status and future directions. Prec Radiat Oncol. 2022; 6:164–176

work page 2022

[4] [4]

Process Modeling a Radiation Oncology Clinic Workflow From Therapeutic Simulation to Treatment: Identifying Impending Strain and Possible Treatment Delays

Chowdhry VK, Simpson NC. Process Modeling a Radiation Oncology Clinic Workflow From Therapeutic Simulation to Treatment: Identifying Impending Strain and Possible Treatment Delays. Adv Radiat Oncol. 2023; 8(6):101261

work page 2023

[5] [5]

Complexity and Approximation - Combinatorial Optimization Problems and Their Approximability Properties 2nd Ed

Ausiello G, Crescenzi P, Gambosi G, Kann V, Marchetti-Spaccamela A, Protasi M. Complexity and Approximation - Combinatorial Optimization Problems and Their Approximability Properties 2nd Ed. Berlin: Springer; 2003

work page 2003

[6] [6]

In-gantry or remote patient positioning? Monte Carlo simulations for proton therapy centers of different sizes

Fava G, et al. In-gantry or remote patient positioning? Monte Carlo simulations for proton therapy centers of different sizes. Radiotherapy and Oncology. 2012; 103:18-24

work page 2012

[7] [7]

Modeling of daily operation in proton radiotherapy by Monte Carlo method

Sakae T, Tsunashima Y, Terunuma T, Sato M. Modeling of daily operation in proton radiotherapy by Monte Carlo method. Jpn J Med Phys. 2003; 23(2):147-156. Japanese

work page 2003

[8] [8]

Healthcare scheduling in 15 optimization context: a review

Abdalkareem ZA, Amir A, Al-Betar MA, Ekhan P, Hammouri AI. Healthcare scheduling in 15 optimization context: a review. Health Technol. 2021; 11:445-469

work page 2021