{"id":"957946eb-7ebd-4422-9764-cf7eeed5583f","arxiv_id":"2412.17813","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A comprehensive review of non-invasive and contactless dehydration monitoring technologies, comparing eight sensing modalities and identifying the lack of gold-standard-annotated datasets as the key bottleneck.","lead":"This paper surveys research on non-invasive and non-contact ways to monitor dehydration, using wearable and remote sensors like smartwatches, radio signals, cameras, and ultrasound. It maps the benefits and limits of eight sensing modalities, catalogs public and private datasets, and argues that future work should combine multiple sensors and collect medically annotated data.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The survey's negative claims of firstness rest on a citation-biased search; a single overlooked non-contact osmolality study would invalidate the central novelty.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing point: the completeness and unbiasedness of the literature search. I agree. The central contribution of this survey is not a new sensing result but a map of the field plus a set of 'first' and 'absence' claims. The authors make universal negative statements ('no work in open literature', 'none of the existing review articles', 'there do not exist medically annotated public datasets') after describing a search that explicitly favors highly cited works and provides no systematic protocol. Because the paper is a review, its value is tied to whether the map is complete; any missed prior work on contactless dehydration sensing or on estimating plasma osmolality from non-contact signals would remove the stated novelty and weaken the research agenda. This is not a matter of disagreeing with the mainstream consensus; it is an internal evidentiary gap between the strength of the claims and the documented method. I also note secondary issues, such as reliance on unpublished self-citations for some performance figures, but those do not change the principal concern. A reproducible, unfiltered systematic search would settle the question. The paper is otherwise informative and well organized, so the appropriate response is to keep the reader's CONDITIONAL verdict: the authors should either moderate the negative claims or provide the missing systematic evidence. No shift in the verdict is needed.","tokens_in":30402,"tokens_out":3231,"duration_ms":32576,"concrete_test":"Run a formal PRISMA-style systematic search across Scopus, Web of Science, PubMed, IEEE Xplore, arXiv, and Google Scholar for 2003-2024 without citation-count or language filters, using combinations such as (dehydration OR hydration) AND (osmolality OR 'plasma osmolality' OR 'serum osmolality') AND (non-contact OR contactless OR camera OR radar OR microwave OR optical OR thermal OR 'remote sensing'), plus a parallel sweep of PhysioNet, MIMIC, eICU, and medical imaging repositories for datasets pairing non-contact sensor data with biochemistry reports. Record query strings, screening decisions, and retrieved studies. If any retrieved study already infers plasma/serum osmolality from non-contact signals, or any existing review covers contactless dehydration sensing or annotated datasets, the paper should soften or retract the corresponding negative claims.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section I's 'Literature review criterion' states that the search 'focused particularly on highly cited and influential works' and gives no inclusion/exclusion protocol, no reproducible query set, and no screening method. This is the load-bearing support for the paper's strongest negative claims: that no existing review covers contactless dehydration sensing, that no medically annotated public datasets exist, and that to date no open-literature work infers plasma/serum osmolality non-invasively and non-contactly. These are universal existence claims, so they require an exhaustive search; a search biased toward highly cited works cannot establish them. The vulnerability is concrete: non-contact RF, optical, and camera-based dehydration methods are fastest-moving around 2023-2024, exactly where a citation-bias filter is most likely to drop recent, lower-cited, or non-indexed works. If any such work regresses RF, optical, thermal, or camera features to serum/plasma osmolality, or any published review already discusses contactless methods, the paper's central novelty and its proposed research agenda lose their foundation. The datasets section is likewise based on a 'quick summary' (Table III) and does not document a systematic sweep of repositories, so the absence of medically annotated public datasets is asserted rather than demonstrated. The concern is about methodology and evidence, not about the authors' intent.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper is a narrative review of non-invasive and non-contact dehydration monitoring in the context of the Internet of medical things (IoMT). It covers the pathophysiology of dehydration, clinical gold-standard methods, eight sensing modalities (fluid-based, BIA, EDA, ECG, acoustic, RF, optical, and thermal), multi-modal sensing, organ-specific and population-specific hydration assessment, available datasets, and a research agenda. The paper claims to be the most comprehensive and up-to-date survey of dehydration monitoring, to be the first to cover contactless sensing methods, and it asserts that no medically annotated public datasets exist and that no work in the open literature infers plasma/serum osmolality in a non-invasive and non-contact manner.","tokens_in":30615,"tokens_out":6028,"duration_ms":52835,"significance":"If the survey's coverage and, especially, its negative claims are accurate, the paper would provide a valuable map of a rapidly developing field and a coherent justification for a research agenda centered on multi-modal sensing and medically annotated datasets. The paper has concrete strengths: a broad modality taxonomy, explicit comparison tables (Tables IV and V), a dataset summary (Table III), a timeline of prior reviews (Figure 4), and an unusually candid discussion of clinical-translation barriers such as the lack of clinical trials and the difficulty of measuring euhydration. These features make the survey useful even where its universal claims are not yet established. However, the paper's novelty and proposed agenda rest on absence claims that are not supported by the documented search method, and several performance figures come from unpublished self-citations; both issues are correctable in revision.","major_comments":[{"comment":"The universal negative claims in Section XIV ('to date there exists no work in open literature that aims to infer them ...') and Section XV ('there do not exist medically annotated public datasets') are not supported by the search protocol described in Section I. The protocol names databases and keywords but gives no inclusion/exclusion criteria, no query strings, no screening method, and explicitly states that the search 'focused particularly on highly cited and influential works.' A citation-biased narrative search cannot establish the absence of a class of works, especially in a fast-moving area where recent, lower-cited, or non-indexed papers are likely to be missed. Because these claims frame the paper's novelty and proposed research agenda, they should either be backed by a systematic and reproducible search (e.g., PRISMA-style with search dates, query strings, and screening logs) or rephrased as 'to the best of our knowledge within the searched corpus,' with the corpus explicitly delimited.","section":"Section I (Literature review criterion), Section XIV, Section XV"},{"comment":"The dataset-related absence claim rests on Table III, which is explicitly a 'quick summary' of 13 entries. Several entries are marked 'On request,' and the table includes works that are not public datasets (e.g., [50] describes a sensor study with six subjects, not a data release). The text does not document a systematic sweep of data repositories such as PhysioNet, Zenodo, IEEE DataPort, or Mendeley Data, nor does it give search dates or dataset inclusion criteria. Without this, the statement that no medically annotated public datasets exist is an assertion rather than a demonstrated result. The table should be expanded into a documented dataset search, and the claim should be qualified accordingly.","section":"Section XIV, Table III"},{"comment":"The claim in Section II.C and Figure 3 that all non-biochemistry methods are 'good at dehydration classification only, and are not capable of doing regression, say, estimation of plasma osmolality' is internally inconsistent with works cited elsewhere in the paper. Section XII reports that Suryadevara et al. predict total body water loss with a mean squared error of 2% [132], and that Ring et al. improve the inference of water loss estimation [58]; these are regression tasks. The sentence should distinguish between regression of a continuous hydration metric such as total body water loss and regression of plasma/serum osmolality specifically, or it should be corrected. As written, it overstates the limitation and weakens the later gap claim.","section":"Section II.C, Figure 3, Section XII"},{"comment":"Several specific accuracy figures are attributed to works by the same author group that are unpublished or available only as preprints: the 92% and 87% accuracies for the LC-resonance skin-capacitance method in Section IV (Ref. [60]), the 98.73% ECG classification accuracy in Section VI (Ref. [94]), and the >90% video-PPG accuracy in Section IX (Ref. [119]). Because a survey's comparative value depends on the reliability of reported performance numbers, these citations should be flagged as unpublished or preprint, accompanied by peer-reviewed versions if available, and should not be presented on the same footing as established published benchmarks.","section":"Sections IV, VI, IX; Refs. [60], [94], [119]"},{"comment":"The claim that 'none of the existing review articles have talked about contactless dehydration sensing methods' is supported only by the authors' characterization in Table V, without a documented protocol for how prior reviews were selected and assessed. Since Table V itself lists a review on wireless body area networks (Ref. [70]) that covers RF-based hydration sensing, the boundary between 'contactless' and 'RF' must be defined explicitly, and each prior review should be checked against that definition. This claim should be softened or systematically substantiated.","section":"Section I, Section XV, Table V"}],"minor_comments":[{"comment":"The term 'EU-hydrated' should be 'euhydrated'; the unusual capitalization appears twice and should be corrected.","section":"Sections II.A and XVI"},{"comment":"The phrase 'fetal implications' should be 'fatal implications.'","section":"Section XVII"},{"comment":"The word 'Unviersity' in the thesis title should be 'University.'","section":"Reference [110]"},{"comment":"The units 'MOSM/KG' and 'M EQ/L' should be typeset consistently as 'mOsm/kg' and 'mEq/L.'","section":"Table I"},{"comment":"The frequency range '1 KHz - 1 MHz' should use lowercase 'kHz' for consistency with standard notation.","section":"Section IV"},{"comment":"The term 'RF-IDentification' should be 'RFID' for consistency with the rest of the text.","section":"Section VIII"},{"comment":"There is a mismatched parenthesis in the sentence 'Table III) also reveals that ...'; the opening 'Table III' should not be followed by a closing parenthesis.","section":"Section XIV"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's main contribution is framed by negative existence claims that the current search method cannot support. I do not see this as an irreparable flaw: the survey's taxonomy and discussion are useful, and the authors can address the issue by adding a systematic search protocol and qualifying the claims. The reliance on three unpublished self-citations for specific accuracy numbers should be disclosed clearly and, if possible, replaced by peer-reviewed references. Please also ask the authors to correct the several typographical and reference-metadata issues listed in the minor comments."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis review is worth reading if you work anywhere near IoMT or hydration sensing. It does a solid job organizing a fragmented literature across eight sensing modalities, and its comparison tables and dataset summary are genuinely handy. The section on contactless methods (RF, optical, acoustic, camera) is the most up-to-date I've seen in one place, and the discussion of why multi-modal sensing and medically annotated datasets are the bottleneck is on target.\n\nThe soft spot is exactly what the stress-test flagged. The paper's strongest claims—that no prior review covers contactless dehydration sensing, that no medically annotated public datasets exist, and that no open-literature work infers plasma/serum osmolality non-invasively—are universal negatives, but the search strategy is described as 'focused particularly on highly cited and influential works' with no inclusion/exclusion protocol or reproducible query set. That's not enough to support those claims. It's a methodological weakness, not a sign of bad faith. The authors do hedge with 'to the best of our knowledge,' but the claims still appear as statements of fact in the abstract and conclusions.\n\nAlso minor: a few performance numbers come from the authors' own unpublished preprints (Refs [60], [94], [119]). That's not disqualifying—self-citation is fine when the work is relevant—but in a review, using unpublished numbers without noting their status is a bit uneven.\n\nThe rest holds up. The physiology primer is accurate, the modality-by-modality discussion is balanced (they include negative results, e.g., the null ECG study), and the concluding research agenda is sensible. The paper doesn't break new ground, but it's a legitimate consolidation.\n\nI'd send it to peer review. The main request would be to either temper the absence claims or add a documented systematic search (queries, databases, screening) so readers can judge coverage. With that, it's a useful reference for students and engineers entering the field.\n\nVerdict: a real contributor, not a breakthrough. Worth refereeing.","headline":"A genuinely useful review of dehydration monitoring that overreaches on 'first to cover contactless' because the literature search isn't systematic.","tokens_in":31172,"tokens_out":2174,"would_cite":true,"duration_ms":20862,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This survey maps non-invasive and non-contact dehydration monitoring across eight sensing modalities and argues the field's next step is multi-modal wearable devices, medically annotated datasets, and clinical trials.","keywords":["dehydration monitoring","non-invasive sensing","non-contact sensing","internet of medical things","machine learning","plasma osmolality","multi-modal sensing","biochemical biomarkers"],"falsifier":"A targeted search for peer-reviewed studies from 2020 to 2025 reporting non-invasive or contactless estimation of plasma or serum osmolality with paired biochemical validation, or the discovery of any public dataset containing non-invasive or non-contact sensor data from hospital dehydration patients alongside blood and urine biochemistry reports, would contradict the paper's central gap claims.","tokens_in":30226,"feed_emoji":"💧","tokens_out":5156,"duration_ms":44859,"temperature":0.7,"pith_summary":"The paper is a methodological review of non-invasive and non-contact dehydration monitoring technologies, presented as the most comprehensive and up-to-date map of the field. It catalogs eight sensing modalities, from sweat and urine biomarkers to bioimpedance, electrocardiography, acoustic, radio-frequency, optical, and thermal sensing, and identifies which have contactless variants. Its central gap claims are that no existing review covers contactless methods, that no public dataset pairs non-invasive sensor data with hospital biochemistry reports, and that no published work infers plasma or serum osmolality in a non-invasive, non-contact way. If these claims are right, the review gives researchers a reliable map of what has been tried and a concrete agenda built on multi-modal edge-AI wearables, medically annotated datasets, and clinical validation.","feed_headline":"Survey maps dehydration monitoring without needles or contact","feed_subtitle":"Classifying hydration exists; predicting blood osmolality without contact does not — no hospital-grade dataset either.","key_machinery":"The survey's organizing instrument is a sensing-modality taxonomy: eight categories, namely fluid-based, bioelectrical impedance analysis, electrodermal activity, electrocardiography, acoustic, radio-frequency, optical, and thermal sensing, together with a separate non-contact analysis and a dataset inventory table. This taxonomy lets the authors derive their cross-cutting claims that no single modality suffices, that almost all modalities have low sensitivity to subtle dehydration, and that the existing literature stops at classification and never reaches regression of plasma osmolality.","core_discovery":"On its own terms, the paper establishes a state-of-the-art map: eight sensing modalities, each examined through its biomarker or physical principle, its reported accuracy where available, and its stated limitations, plus a dedicated section on non-contact methods using radio-frequency, acoustic, optical, and camera-based signals. It also compiles a dataset inventory showing that publicly available datasets are small, lab-scale, and largely unannotated by biochemistry, and asserts that no open-literature work has yet inferred plasma or serum osmolality in a non-invasive and non-contact manner. The authors conclude that a single modality cannot give the full picture of hydration status, so the future lies in combining sensors in IoMT wearable devices with machine learning, in collecting medically annotated datasets, and in running clinical trials to earn clinician trust.","pith_inferences":["A public challenge dataset pairing contactless RF, optical, or thermal signals with paired plasma osmolality readings would directly test whether the regression gap is a data problem or a fundamental signal-information problem.","If osmolality regression proves feasible from one or more contactless modalities, mass screening at events such as marathons or construction sites becomes plausible, though the survey's own cautions about inter-subject variability suggest per-subject calibration will likely be needed first.","The same dataset scarcity may explain why the field clusters on classification; generative synthetic data could fill gaps only after real medically annotated examples exist, so the near-term bottleneck is data collection rather than model innovation.","The boundary between 'non-invasive' and 'non-contact' may blur as wearables become thinner, and future work could re-examine which claimed contactless methods are truly unobtrusive versus minimal-contact."],"forward_implications":["Researchers get a consolidated map of which sensing modalities have demonstrated contactless variants and which remain contact-based, lowering the cost of entry for new work.","The absence of public medically annotated datasets becomes a precisely defined bottleneck: raw sensor data paired with blood and urine biochemistry from hospital dehydration patients is the missing resource.","The 'classification only, no regression' pattern gives a concrete target: the next milestone is a non-invasive, non-contact estimator of plasma or serum osmolality, not another hydrated-versus-dehydrated classifier.","Multi-modal fusion is positioned as the route to overcome single-modality limitations and inter-subject variability, with edge-AI wearable devices as the envisioned deployment form.","The need for clinical trials is stated as the condition for clinician trust and eventual adoption of wearable and contactless methods."],"supporting_citations":[{"why":"Most recent prior review of non-invasive hydration assessment, restricted to urolithiasis patients, which the survey uses to position its own broader scope.","marker":"[13]"},{"why":"Athlete-focused hydration review that represents the narrow-population reviews the survey contrasts with its comprehensive coverage.","marker":"[14]"},{"why":"Establishes plasma osmolality as the most reputed dehydration biomarker, the gold standard that the survey claims no non-invasive and non-contact work has inferred.","marker":"[17]"},{"why":"Supplies the physiological basis of dehydration assessment and its effects on performance, used to frame the clinical importance of monitoring.","marker":"[18]"},{"why":"Provides practical measures and empirical thresholds that improve fluid-based hydration assessment, forming the basis of the survey's discussion of fluid-method limitations.","marker":"[49]"},{"why":"A 2023 review of noninvasive dehydration monitoring that the survey claims did not discuss non-contact techniques, supporting the novelty claim.","marker":"[79]"},{"why":"Public smartphone video-PPG dehydration dataset by the same group, one of the few public datasets listed in the inventory.","marker":"[119]"},{"why":"Public wearable GSR and PPG dataset used in the dataset table, evidence for the claim that public datasets exist but are not medically annotated.","marker":"[139]"}],"fun_headline_variants":["Dehydration detection without needles or contact: survey","No-contact dehydration monitoring: challenges and gaps","Contactless dehydration sensing: the missing dataset","IoMT hydration monitors: no contact, but no proof yet","Survey: contactless dehydration detection still unproven"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The survey's negative claims rest on the completeness and unbiasedness of a narrative literature search that explicitly focused on highly cited and influential works, so a missed recent or less-cited study on contactless sensing or on non-invasive osmolality estimation would weaken the paper's central novelty and its proposed research agenda.","fun_headline_variants_meta":{"raw":{"variants":["Dehydration detection without needles or contact: survey","No-contact dehydration monitoring: challenges and gaps","Contactless dehydration sensing: the missing dataset","IoMT hydration monitors: no contact, but no proof yet","Survey: contactless dehydration detection still unproven"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000544,"raw_usage":{"total_tokens":2595,"prompt_tokens":925,"completion_tokens":1670,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":1611}},"tokens_in":541,"tokens_out":1670,"duration_ms":13049,"temperature":1.0,"reasoning_tokens":1611,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:56:32.377022+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A targeted search for peer-reviewed studies from 2020 to 2025 reporting non-invasive or contactless estimation of plasma or serum osmolality with paired biochemical validation, or the discovery of any public dataset containing non-invasive or non-contact sensor data from hospital dehydration patients alongside blood and urine biochemistry reports, would contradict the paper's central gap claims.","supporting_citations":[{"cited_title":"Noninvasive monitoring to detect dehydration: Are we there yet?","cited_arxiv_id":null,"evidence_quote":"A 2023 review of noninvasive dehydration monitoring that the survey claims did not discuss non-contact techniques, supporting the novelty claim."},{"cited_title":"You can monitor your hydration level using your smartphone camera","cited_arxiv_id":"2402.07467","evidence_quote":"Public smartphone video-PPG dehydration dataset by the same group, one of the few public datasets listed in the inventory."},{"cited_title":"Towards on-device dehydration monitoring using machine learning from wearable device’s data,","cited_arxiv_id":null,"evidence_quote":"Public wearable GSR and PPG dataset used in the dataset table, evidence for the claim that public datasets exist but are not medically annotated."}],"review_version":1}