Segmentation metric misinterpretations in bioimage analysis.

Quantitative evaluation of image segmentation algorithms is crucial in the field of bioimage analysis. The most common assessment scores, however, are often misinterpreted and multiple definitions coexist with the same name. Here we present the ambiguities of evaluation metrics for segmentation algorithms and show how these misinterpretations can alter leaderboards of influential competitions. We also propose guidelines for how the currently existing problems could be tackled.

Three million images and morphological profiles of cells treated with matched chemical and genetic perturbations.

The identification of genetic and chemical perturbations with similar impacts on cell morphology can elucidate compounds' mechanisms of action or novel regulators of genetic pathways. Research on methods for identifying such similarities has lagged due to a lack of carefully designed and well-annotated image sets of cells treated with chemical and genetic perturbations. Here we create such a Resource dataset, CPJUMP1, in which each perturbed gene's product is a known target of at least two chemical compounds in the dataset. We systematically explore the directionality of correlations among perturbations that target the same protein encoded by a given gene, and we find that identifying matches between chemical and genetic perturbations is a challenging task. Our dataset and baseline analyses provide a benchmark for evaluating methods that measure perturbation similarities and impact, and more generally, learn effective representations of cellular state from microscopy images. Such advancements would accelerate the applications of image-based profiling of cellular states, such as uncovering drug mode of action or probing functional genomics.

High-dimensional gene expression and morphology profiles of cells across 28,000 genetic and chemical perturbations.

Cells can be perturbed by various chemical and genetic treatments and the impact on gene expression and morphology can be measured via transcriptomic profiling and image-based assays, respectively. The patterns observed in these high-dimensional profile data can power a dozen applications in drug discovery and basic biology research, but both types of profiles are rarely available for large-scale experiments. Here, we provide a collection of four datasets with both gene expression and morphological profile data useful for developing and testing multimodal methodologies. Roughly a thousand features are measured for each of the two data types, across more than 28,000 chemical and genetic perturbations. We define biological problems that use the shared and complementary information in these two data modalities, provide baseline analysis and evaluation metrics for multi-omic applications, and make the data resource publicly available ( https://broad.io/rosetta/ ).

Hereditary thrombotic thrombocytopenic purpura acquired through liver transplantation: A case report.

A 62-year-old received orthotopic liver transplantation. Three weeks later, thrombotic microangiopathy developed. Testing revealed thrombotic thrombocytopenic purpura (TTP) characterized by low ADAMTS13 (A Disintegrin-like Metallopeptidase with ThromboSpondin type 1 motif 13) activity and no inhibitor of ADAMTS13 protein. Retrospective attainment of donor records revealed a TTP diagnosis, presumably hereditary TTP (hTTP), as an ADAMTS13 protein inhibitor was not mentioned. As the grafted liver does not produce ADAMTS13 protein, the recipient now functionally has hTTP and will likely need plasma transfusions indefinitely. While hTTP is extremely rare, it should be considered a contraindication to liver donation outside of exceptional circumstances. If a potential liver donor has TTP listed on medical history, attempts should be made to determine whether it is autoimmune or hereditary. An accurate medical history is critical as it is the only reliable way to identify hTTP, as outside of acute exacerbations of TTP, donors with hTTP can have normal laboratory values, including normal hemoglobin, platelets, and renal function.

Publisher Correction: Nucleus segmentation across imaging experiments: the 2018 Data Science Bowl.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Anxiety and depression in association with lower urinary tract symptoms: results from the COBaLT study.

PURPOSE: Depression and anxiety have been associated with lower urinary tract symptoms (LUTS) in several studies. In our population, the prevalence of LUTS is high, consequently, it is essential and an objective of this study, to determine the association between anxiety, depression, and LUTS in a large Hispanic population. METHODS: A sub-analysis of a cross-sectional population-based study to estimate LUTS prevalence in the Colombian population was performed (COBaLT study) (Plata et al. in Neurourol Urodyn 38:200-207, 2018). The Hospital Anxiety and Depression Scale (HADS) was used to evaluate mental health. Logistic regression was carried out to estimate the association of depression and anxiety with different LUTS. Variables that proved statistically significant (p < 0.05) were included in a multivariate model. RESULTS: A total of 1060 individuals were assessed. The prevalence of anxiety in women and men was 17.1% and 6.7%, respectively. Depression in women and men was 20.1% and 9.4%, respectively. An association was found between anxiety and overactive bladder (OAB) without urinary incontinence (OR = 3.7) and moderate or severe LUTS in men (OR = 3.8). In women, anxiety was associated with nocturia (OR = 4.2) and stress urinary incontinence (OR = 2.4). For depression, an association was found between sexual dysfunction (OR = 4.3) and moderate or severe LUTS (OR = 4.0) in men; while in women it was associated with stress urinary incontinence (OR = 2.3), the sensation of incomplete emptying (OR = 1.7) and decreased frequency of sexual activity (OR = 1.8). CONCLUSIONS: The associations found are consistent with other reports. It is essential to inquire about possible symptoms related to the mental sphere in the urology consultation to make appropriate referrals and subsequent management.

Objective assessment of stored blood quality by deep learning.

Stored red blood cells (RBCs) are needed for life-saving blood transfusions, but they undergo continuous degradation. RBC storage lesions are often assessed by microscopic examination or biochemical and biophysical assays, which are complex, time-consuming, and destructive to fragile cells. Here we demonstrate the use of label-free imaging flow cytometry and deep learning to characterize RBC lesions. Using brightfield images, a trained neural network achieved 76.7% agreement with experts in classifying seven clinically relevant RBC morphologies associated with storage lesions, comparable to 82.5% agreement between different experts. Given that human observation and classification may not optimally discern RBC quality, we went further and eliminated subjective human annotation in the training step by training a weakly supervised neural network using only storage duration times. The feature space extracted by this network revealed a chronological progression of morphological changes that better predicted blood quality, as measured by physiological hemolytic assay readouts, than the conventional expert-assessed morphology classification system. With further training and clinical testing across multiple sites, protocols, and instruments, deep learning and label-free imaging flow cytometry might be used to routinely and objectively assess RBC storage lesions. This would automate a complex protocol, minimize laboratory sample handling and preparation, and reduce the impact of procedural errors and discrepancies between facilities and blood donors. The chronology-based machine-learning approach may also improve upon humans' assessment of morphological changes in other biomedically important progressions, such as differentiation and metastasis.

Analysis of Diversity, Equity, and Inclusion Initiatives Presented Across US Academic Department of Surgery Websites in 2021.

Despite increased attention devoted to diversity, equity, and inclusion (DEI) within academic medicine, representation, lack of workforce and leadership diversity, and bias within medicine remain persistent problems. The purpose of the current study was to understand the current efforts and attention to DEI within academic departments of surgery in the United States. 251 department of surgery websites were reviewed, using a standardized data collection form and scoring procedure, accompanied by a 10 percent fidelity check by an independent reviewer. Only 16% of departments of surgery included DEI-specific information, such as a DEI mission statement or initiatives on their departmental sites, with less than seven percent of departments reporting a DEI committee. Such public information may have implications for recruitment and retention of diverse faculty and trainees, downstream effects for patient care, and could be critical to public accountability to improve diversity and create a culture of equity and inclusion.

Probabilistic detection of impacts using the PFEEL algorithm with a Gaussian Process Regression Model.

Methods for identifying human activity have a wide range of potential applications, including security, event time detection, intelligent building environments, and human health. Current methodologies typically rely on either wave propagation or structural dynamics principles. However, force-based methods, such as the probabilistic force estimation and event localization algorithm (PFEEL), offer advantages over wave propagation methods by avoiding challenges such as multi-path fading. PFEEL utilizes a probabilistic framework to estimate the force of impacts and the event locations in the calibration space, providing a measure of uncertainty in the estimations. This paper presents a new implementation of PFEEL using a data-driven model based on Gaussian process regression (GPR). The new approach was evaluated using experimental data collected on an aluminum plate impacted at eighty-one points, with a separation of five centimeters. The results are presented as an area of localization relative to the actual impact location at different probability levels. These results can aid analysts in determining the required precision for various implementations of PFEEL.

Impact of having potential living donors on ethnic/racial disparities in access to kidney transplantation.

Racial/ethnic disparities persist in patients' access to living donor kidney transplantation (LDKT). This study assessed the impact of having available potential living donors (PLDs) on candidates' receipt of a kidney transplant (KT) and LDKT at two KT programs. Using data from our clinical trial of waitlisted candidates (January 1, 2014-December 31, 2019), we evaluated Hispanic and Non-Hispanic White (NHW) KT candidates' number of PLDs. Multivariable logistic regression assessed the impact of PLDs on transplantation (KT vs. no KT; for KT recipients, LDKT vs. deceased donor KT). A total of 847 candidates were included, identifying as Hispanic (45.8%) or NHW (54.2%). For Site A, both Hispanic (adjusted OR = 2.26 [95% CI 1.13-4.53]) and NHW (OR = 2.42 [1.10-5.33]) candidates with PLDs completing the questionnaire were more likely to receive a KT. For Site B, candidates with PLDs were not significantly more likely to receive KT. Among KT recipients at both sites, Hispanic (Site A: OR = 21.22 [2.44-184.88]; Site B: OR = 25.54 [7.52-101.54]), and NHW (Site A: OR = 37.70 [6.59-215.67]; Site B: OR = 15.18 [5.64-40.85]) recipients with PLD(s) were significantly more likely to receive a LDKT. Our findings suggest that PLDs increased candidates' likelihood of KT receipt, particularly LDKT. Transplant programs should help candidates identify PLDs early in transplant evaluation.

Effectiveness of a culturally competent care intervention in reducing disparities in Hispanic live donor kidney transplantation: A hybrid trial.

Hispanic patients receive disproportionately fewer living donor kidney transplants (LDKTs) than non-Hispanic Whites (NHWs). The Northwestern Medicine Hispanic Kidney Transplant Program (HKTP), designed to increase Hispanic LDKTs, was evaluated as a nonrandomized, implementation-effectiveness hybrid trial of patients initiating transplant evaluation at two intervention and two similar control sites. Using a mixed method, observational design, we evaluated the fidelity of the HKTP implementation at the two intervention sites. We tested the impact of the HKTP intervention by evaluating the likelihood of receiving LDKT comparing pre-intervention (January 2011-December 2016) and postintervention (January 2017-March 2020), across ethnicity and centers. The HKTP study included 2063 recipients. Intervention Site A exhibited greater implementation fidelity than intervention Site B. For Hispanic recipients at Site A, the likelihood of receiving LDKTs was significantly higher at postintervention compared with pre-intervention (odds ratio [OR] = 3.17 95% confidence interval [1.04, 9.63]), but not at the paired control Site C (OR = 1.02 [0.61, 1.71]). For Hispanic recipients at Site B, the likelihood of receiving an LDKT did not differ between pre- and postintervention (OR = 0.88 [0.40, 1.94]). The LDKT rate was significantly lower for Hispanics at paired control Site D (OR = 0.45 [0.28, 0.90]). The intervention significantly improved LDKT rates for Hispanic patients at the intervention site that implemented the intervention with greater fidelity. Registration: ClinicalTrials.gov registered (retrospectively) on September 7, 2017 (NCT03276390).

A multicentre outcome analysis to define global benchmarks for donation after circulatory death liver transplantation.

BACKGROUND & AIMS: The concept of benchmarking is established in the field of transplant surgery; however, benchmark values for donation after circulatory death (DCD) liver transplantation are not available. Thus, we aimed to identify the best possible outcomes in DCD liver transplantation and to propose outcome reference values. METHODS: Based on 2,219 controlled DCD liver transplantations, collected from 17 centres in North America and Europe, we identified 1,012 low-risk, primary, adult liver transplantations with a laboratory MELD score of ≤20 points, receiving a DCD liver with a total donor warm ischemia time of ≤30 minutes and asystolic donor warm ischemia time of ≤15 minutes. Clinically relevant outcomes were selected and complications were reported according to the Clavien-Dindo-Grading and the comprehensive complication index (CCI). Corresponding benchmark cut-offs were based on median values of each centre, where the 75th-percentile was considered. RESULTS: Benchmark cases represented between 19.7% and 75% of DCD transplantations in participating centres. The 1-year retransplant and mortality rates were 4.5% and 8.4% in the benchmark group, respectively. Within the first year of follow-up, 51.1% of recipients developed at least 1 major complication (≥Clavien-Dindo-Grade III). Benchmark cut-offs were ≤3 days and ≤16 days for ICU and hospital stay, ≤66% for severe recipient complications (≥Grade III), ≤16.8% for ischemic cholangiopathy, and ≤38.9 CCI points 1 year after transplant. Comparisons with higher risk groups showed more complications and impaired graft survival outside the benchmark cut-offs. Organ perfusion techniques reduced the complications to values below benchmark cut-offs, despite higher graft risk. CONCLUSIONS: Despite excellent 1-year survival, morbidity in benchmark cases remains high. Benchmark cut-offs targeting morbidity parameters offer a valid tool to assess the protective value of new preservation technologies in higher risk groups and to provide a valid comparator cohort for future clinical trials. LAY SUMMARY: The best possible outcomes after liver transplantation of grafts donated after circulatory death (DCD) were defined using the concept of benchmarking. These were based on 2,219 liver transplantations following controlled DCD donation in 17 centres worldwide. Donor and recipient combinations with higher risk had significantly worse outcomes. However, the use of novel organ perfusion technology helped high-risk patients achieve similar outcomes as the benchmark cohort.

Nucleus segmentation across imaging experiments: the 2018 Data Science Bowl.

Segmenting the nuclei of cells in microscopy images is often the first step in the quantitative analysis of imaging data for biological and biomedical applications. Many bioimage analysis tools can segment nuclei in images but need to be selected and configured for every experiment. The 2018 Data Science Bowl attracted 3,891 teams worldwide to make the first attempt to build a segmentation method that could be applied to any two-dimensional light microscopy image of stained nuclei across experiments, with no human interaction. Top participants in the challenge succeeded in this task, developing deep-learning-based models that identified cell nuclei across many image types and experimental conditions without the need to manually adjust segmentation parameters. This represents an important step toward configuration-free bioimage analysis software tools.

Black Patients With Cirrhosis Have Higher Mortality and Lower Transplant Rates: Results From a Metropolitan Cohort Study.

BACKGROUND AND AIMS: Estimates of racial disparity in cirrhosis have been limited by lack of large-scale, longitudinal data, which track patients from diagnosis to death and/or transplant. APPROACH AND RESULTS: We analyzed a large, metropolitan, population-based electronic health record data set from seven large health systems linked to the state death registry and the national transplant database. Multivariate competing risk analyses, adjusted for sex, age, insurance status, Elixhauser score, etiology of cirrhosis, HCC, portal hypertensive complication, and Model for End-Stage Liver Disease-Sodium (MELD-Na), examined the relationship between race, transplant, and cause of death as defined by blinded death certificate review. During the study period, 11,277 patients met inclusion criteria, of whom 2,498 (22.2%) identified as Black. Compared to White patients, Black patients had similar age, sex, MELD-Na, and proportion of alcohol-associated liver disease, but higher comorbidity burden, lower rates of private insurance, and lower rates of portal hypertensive complications. Compared to White patients, Black patients had the highest rate all-cause mortality and non-liver-related death and were less likely to be listed or transplanted (P < 0.001 for all). In multivariate competing risk analysis, Black patients had a 26% increased hazard of liver-related death (subdistribution HR, 1.26; 95% CI, [1.15-1.38]; P < 0.001). CONCLUSIONS: Black patients with cirrhosis have discordant outcomes. Further research is needed to determine how to address these real disparities in the field of hepatology.

Liver Transplantation Following Yttrium-90 Radioembolization: 15-Year Experience in 207-Patient Cohort.

BACKGROUND AND AIMS: Radioembolization (yttrium-90 [Y90]) is used in hepatocellular carcinoma (HCC) as a bridging as well as downstaging liver-directed therapy to curative liver transplantation (LT). In this study, we report long-term outcomes of LT for patients with HCC who were bridged/downstaged by Y90. APPROACH AND RESULTS: Patients undergoing LT following Y90 between 2004 and 2018 were included, with staging by United Network for Organ Sharing (UNOS) tumor-node-metastasis criteria at baseline pre-Y90 and pre-LT. Post-Y90 toxicities were recorded. Histopathological data of HCC at explant were recorded. Long-term outcomes, including overall survival (OS), recurrence-free survival (RFS), disease-specific mortality (DSM), and time-to-recurrence, were reported. Time-to-endpoint analyses were estimated using Kaplan-Meier. Univariate and multivariate analyses were performed using a log-rank test and Cox proportional-hazards model, respectively. During the 15-year period, 207 patients underwent LT after Y90. OS from LT was 12.5 years, with a median time to LT of 7.5 months [interquartile range, 4.4-10.3]. A total of 169 patients were bridged, whereas 38 were downstaged to LT. Respectively, 94 (45%), 60 (29%), and 53 (26%) patients showed complete, extensive, and partial tumor necrosis on histopathology. Three-year, 5-year, and 10-year OS rates were 84%, 77%, and 60%, respectively. Twenty-four patients developed recurrence, with a median RFS of 120 (95% confidence interval, 69-150) months. DSM at 3, 5, and 10 years was 6%, 11%, and 16%, respectively. There were no differences in OS/RFS for patients who were bridged or downstaged. RFS was higher in patients with complete/extensive versus partial tumor necrosis (P < 0.0001). For patients with UNOS T2 treated during the study period, 5.2% dropped out because of disease progression. CONCLUSIONS: Y90 is an effective treatment for HCC in the setting of bridging/downstaging to LT. Patients who achieved extensive or complete necrosis had better RFS, supporting the practice of neoadjuvant treatment before LT.

Which recipient pretransplant factors, such as MELD, renal function, sarcopenia, and recent sepsis influence suitability for and outcome after living donor liver transplantation? A systematic review of the literature and expert panel recommendations.

BACKGROUND: Varied access to deceased donors across the globe has resulted in differential living donor liver transplant (LDLT) practices and lack of consensus over the influence of models for end stage liver disease (MELD), renal function, sarcopenia, or recent infection on short-term outcomes. OBJECTIVES: Consider these risk factors in relation to patient selection and provide recommendations. DATA SOURCES: Ovid MEDLINE, Embase, Scopus, Google Scholar, Cochrane Central. METHODS: PRIMSA systematic review and GRADE. PROSPERO ID: RD42021260809 RESULTS: MELD >25-30 alone is not a contraindication to LDLT, and multiple studies found no increase in short term mortality in high MELD patients. Contributing factors such as muscle mass, acute physiologic assessment and chronic health evaluation score, donor age, graft weight/recipient weight ratio, and inclusion of the middle hepatic vein in a right lobe graft influence morbidity and mortality in high MELD patients. Higher mortality is observed with pretransplant renal dysfunction, but short-term mortality is rare. Sarcopenia and recent infection are not contraindications to LDLT. Morbidity and prolonged LOS are common, and more frequent in patients with renal dysfunction, nutritional deficiency or recent infection. CONCLUSIONS: When individual risk factors are studied mortality is low and graft loss is infrequent, but morbidity is common. MELD, especially with concomitant risk factors, had the greatest influence on short term outcome, and recent infection had the least. A multidisciplinary team of experts should carefully assess patients with multiple risk factors, and an optimal graft is recommended.

Impact of Portal Hypertension on Adverse Events after Splenic Arterial Aneurysm Embolization.

PURPOSE: To evaluate the outcomes of splenic artery aneurysm (SAA) embolization and compare adverse event (AE) rates after embolization in patients with and without portal hypertension (PHTN). MATERIALS AND METHODS: A retrospective review of all patients who underwent embolization of SAAs at 2 institutions was performed (34 patients from institution 1 and 7 patients from institution 2). Baseline demographic characteristics, preprocedural imaging, procedural techniques, and postprocedural outcomes were evaluated. Thirty-day postprocedural severe and life-threatening AEs were evaluated using the Society of Interventional Radiology guidelines. Thirty-day mortality and readmission rates were also evaluated. t test, χ2 test, and/or Fisher exact test were used for the statistical analysis. RESULTS: There was no statistically significant difference between patients with and without PHTN in the location, number, and size of SAA(s). All procedures were technically successful. There were 13 (32%) patients with and 28 (68%) patients without PHTN. The 30-day mortality rate (31% vs 0%; P = .007), readmission rates (61% vs 7%; P < .001), and severe/life-threatening AE rates (69% vs 0%; P < .001) were significantly higher in patients with PHTN than in those without PHTN. CONCLUSIONS: There was a significantly higher mortality and severe/life-threatening AE rate in patients with PHTN than in those without PHTN. SAAs in patients with PHTN need to be managed very cautiously, given the risk of severe/life-threatening AEs after embolization.

What is the real burden of the overactive bladder? Results from a national prevalence study.

OBJECTIVE: To describe the prevalence of overactive bladder determining patient-reported outcome measures (PROMS) and potential risk factors. METHODS: A cross-sectional population-based study to estimate lower urinary tract symptoms (LUTS) prevalence in the Colombian population was performed (COBaLT study). Overactive bladder (OAB) was assessed using 2002 International Continence Society definitions and the Report on the Terminology for Female Pelvic Floor Dysfunction. PROMS were included to evaluate participant's perceptions of health-related quality of life (QoL). Spanish validated questionnaires were used, including the International Consultation on Incontinence Questionnaire Overactive Bladder (ICIQ-OAB) and Patient Perception of Bladder Condition (PPBC) scale. RESULTS: A total of 1060 individuals were assessed. The mean age of participants was 42.2 years, and most participants were Hispanic (93.15%). The overall prevalence of OAB syndrome was 31.70%, and it was more frequently reported in women than in men (39.25% vs. 24.15%). Most participants with OAB did not perceive their symptoms as bothersome. The most bothersome symptom, associated with moderate/severe impact in QoL was urinary urgency. According to the PPBC questionnaire 75.6% of the participants reported that their bladder condition does not cause any problem, while 11.% reported moderate to severe bother. The multivariable model showed high blood pressure and anxiety were associated with OAB in men. In females, depression, obstructive sleep apnea, IBS, and pelvic organ prolapse were associated with OAB syndrome. CONCLUSIONS: Using the symptomatic definition of OAB can overestimate the real impact and burden of the condition, and treatment should be targeted to those symptomatic patients with QoL impairment as they would benefit from further management. The questionnaires seem to overestimate the condition, since we assessed it from a PROMS perspective there is not such a marked QoL impact.

Gaussian processes retrieval of crop traits in Google Earth Engine based on Sentinel-2 top-of-atmosphere data.

The unprecedented availability of optical satellite data in cloud-based computing platforms, such as Google Earth Engine (GEE), opens new possibilities to develop crop trait retrieval models from the local to the planetary scale. Hybrid retrieval models are of interest to run in these platforms as they combine the advantages of physically- based radiative transfer models (RTM) with the flexibility of machine learning regression algorithms. Previous research with GEE primarily relied on processing bottom-of-atmosphere (BOA) reflectance data, which requires atmospheric correction. In the present study, we implemented hybrid models directly into GEE for processing Sentinel-2 (S2) Level-1C (L1C) top-of-atmosphere (TOA) reflectance data into crop traits. To achieve this, a training dataset was generated using the leaf-canopy RTM PROSAIL in combination with the atmospheric model 6SV. Gaussian process regression (GPR) retrieval models were then established for eight essential crop traits namely leaf chlorophyll content, leaf water content, leaf dry matter content, fractional vegetation cover, leaf area index (LAI), and upscaled leaf variables (i.e., canopy chlorophyll content, canopy water content and canopy dry matter content). An important pre-requisite for implementation into GEE is that the models are sufficiently light in order to facilitate efficient and fast processing. Successful reduction of the training dataset by 78% was achieved using the active learning technique Euclidean distance-based diversity (EBD). With the EBD-GPR models, highly accurate validation results of LAI and upscaled leaf variables were obtained against in situ field data from the validation study site Munich-North-Isar (MNI), with normalized root mean square errors (NRMSE) from 6% to 13%. Using an independent validation dataset of similar crop types (Italian Grosseto test site), the retrieval models showed moderate to good performances for canopy-level variables, with NRMSE ranging from 14% to 50%, but failed for the leaf-level estimates. Obtained maps over the MNI site were further compared against Sentinel-2 Level 2 Prototype Processor (SL2P) vegetation estimates generated from the ESA Sentinels' Application Platform (SNAP) Biophysical Processor, proving high consistency of both retrievals (R 2 from 0.80 to 0.94). Finally, thanks to the seamless GEE processing capability, the TOA-based mapping was applied over the entirety of Germany at 20 m spatial resolution including information about prediction uncertainty. The obtained maps provided confidence of the developed EBD-GPR retrieval models for integration in the GEE framework and national scale mapping from S2-L1C imagery. In summary, the proposed retrieval workflow demonstrates the possibility of routine processing of S2 TOA data into crop traits maps at any place on Earth as required for operational agricultural applications.

Hybrid retrieval of crop traits from multi-temporal PRISMA hyperspectral imagery.

The recently launched and upcoming hyperspectral satellite missions, featuring contiguous visible-to-shortwave infrared spectral information, are opening unprecedented opportunities for the retrieval of a broad set of vegetation traits with enhanced accuracy through novel retrieval schemes. In this framework, we exploited hyperspectral data cubes collected by the new-generation PRecursore IperSpettrale della Missione Applicativa (PRISMA) satellite of the Italian Space Agency to develop and test a hybrid retrieval workflow for crop trait mapping. Crop traits were mapped over an agricultural area in north-east Italy (Jolanda di Savoia, FE) using PRISMA images collected during the 2020 and 2021 vegetative seasons. Leaf chlorophyll content, leaf nitrogen content, leaf water content and the corresponding canopy level traits scaled through leaf area index were estimated using a hybrid retrieval scheme based on PROSAIL-PRO radiative transfer simulations coupled with a Gaussian processes regression algorithm. Active learning algorithms were used to optimise the initial set of simulated data by extracting only the most informative samples. The accuracy of the proposed retrieval scheme was evaluated against a broad ground dataset collected in 2020 in correspondence of three PRISMA overpasses. The results obtained were positive for all the investigated variables. At the leaf level, the highest accuracy was obtained for leaf nitrogen content (LNC: r2=0.87, nRMSE=7.5%), while slightly worse results were achieved for leaf chlorophyll content (LCC: r2=0.67, nRMSE=11.7%) and leaf water content (LWC: r2=0.63, nRMSE=17.1%). At the canopy level, a significantly higher accuracy was observed for nitrogen content (CNC: r2=0.92, nRMSE=5.5%) and chlorophyll content (CCC: r2=0.82, nRMSE=10.2%), whereas comparable results were obtained for water content (CWC: r2=0.61, nRMSE=16%). The developed models were additionally tested against an independent dataset collected in 2021 to evaluate their robustness and exportability. The results obtained (i. e., LCC: r2=0.62, nRMSE=27.9%; LNC: r2=0.35, nRMSE=28.4%; LWC: r2=0.74, nRMSE=20.4%; LAI: r2=0.84, nRMSE=14.5%; CCC: r2=0.79, nRMSE=18.5%; CNC: r2=0.62, nRMSE=23.7%; CWC: r2=0.92, nRMSE=16.6%) evidence the transferability of the hybrid approach optimised through active learning for most of the investigated traits. The developed models were then used to map the spatial and temporal variability of the crop traits from the PRISMA images. The high accuracy and consistency of the results demonstrates the potential of spaceborne imaging spectroscopy for crop monitoring, paving the path towards routine retrievals of multiple crop traits over large areas that could drive more effective and sustainable agricultural practices worldwide.