Validation of a Body-Composition Segmentation Software on a Diverse Public CT Scan Cohort
Trial statusNot yet recruiting
Trial phaseNot listed
Trial typeObservational
Biological sexAll
Age18+
SponsorNucleo Research, Inc.
About this trial
This study evaluates the standalone performance of Soma, a deep-learning software developed by Nucleo Research, Inc. for the automated segmentation of body-composition tissues (skeletal muscle, subcutaneous adipose tissue, visceral adipose tissue, and intramuscular adipose tissue) on whole-body computed tomography (CT) images. The aim is to confirm that Soma produces segmentations and tissue-area measurements that agree with a multi-rater expert reference standard, on a diverse cohort representative of demographic and clinical variation. A total of 200 CT scans are sampled by stratified design from a curated pool of 2,066 scans aggregated from six publicly available, de-identified imaging datasets (autoPET, AMOS, MSD Pancreas, CT-ORG, ENHANCE.PET, RATIC). Three board-certified radiologists independently annotate the reference standard at the L3 slice. Primary performance is assessed using the Dice similarity coefficient against the multi-rater reference, with predefined thresholds and BCa bootstrap confidence intervals, both in aggregate and within every demographic and clinical subgroup. Secondary endpoints include Bland-Altman analysis of tissue-area agreement, 95th-percentile Hausdorff distance, Pearson correlation of derived indices, and Cohen's kappa for sarcopenia classification using Skeletal Muscle Index (SMI). The study is fully retrospective on de-identified images, involves no patient contact, and has been determined exempt by Salus IRB (Salus Number 26328) under 45 CFR 46.104(d)(4).
Eligibility criteria
This trial accepts healthy volunteersQualifiers
Subjects above 16 years or older at the time the source imaging was acquired.
De-identified abdominal computed tomography (CT) scan available from one of the six predefined publicly available datasets (autoPET, AMOS, MSD Pancreas, CT-ORG, ENHANCE.PET, or RATIC).
Scan covers the third lumbar vertebra (L3) with a contiguous axial slice suitable for L3-level body-composition analysis.
Demographic metadata required for stratified sampling (age, sex; BMI where available; clinical context as encoded in source dataset) is present.
Disqualifiers
Subject under 16 years of age at the time the source imaging was acquired.
Scan does not include the L3 vertebra or has severe motion artifact, truncation, or metallic artifact precluding analysis at the L3 level.
Duplicate or near-duplicate scans of the same subject already included in the cohort.
Missing demographic metadata required for at least one stratification axis.
Trial population
Subjects older than 16 years old whose de-identified abdominal CT imaging is available from one of six publicly accessible imaging datasets (autoPET, AMOS, MSD Pancreas, CT-ORG, ENHANCE.PET, RATIC). The 200-scan study cohort is selected by stratified sampling from a curated pool of 2,066 scans to ensure representation across BMI category, age band, sex, body region (abdomen-only vs. whole-body), and clinical context (oncologic vs. non-oncologic).
Trial design
Design model
Cohort
Time perspective
Retrospective
Treatments tested in this trial
Soma Body-Composition Segmentation Software
Diagnostic testSoma is a deep-learning software pipeline developed by Nucleo Research, Inc. for the automated quantitative analysis of body composition from abdominal CT. It comprises (i) a U-Net segmentation model that delineates skeletal muscle, subcutaneous adipose tissue, visceral adipose tissue, and intramuscular adipose tissue on each axial CT slice; and (ii) an EfficientNet-Lite0 + BiLSTM model for automated L3 vertebra detection from axial CT volumes. In this validation study, segmentation performance is assessed on every fifth axial slice across the full scan depth. Outputs include per-tissue segmentation masks, tissue cross-sectional areas (cm\^2), and derived indices including the Skeletal Muscle Index (SMI = muscle area / height\^2). In this study, Soma is applied as the index test in standalone mode, fully blinded to the multi-rater radiologist reference standard.
Treatment groups
200 Participants
are divided into 1 treatment groupGroup A: Public CT Validation Cohort1 intervention
Trial outcomes
Primary outcomes
1
Dice Similarity Coefficient (DSC) of Soma Segmentation Versus Multi-Rater Radiologist Reference Standard
Mean Dice Similarity Coefficient (DSC) between Soma-generated segmentation masks and the consensus reference from three board-certified radiologists, computed per tissue class (skeletal muscle, subcutaneous adipose tissue, visceral adipose tissue, intramuscular adipose tissue) on all annotated axial slices (every fifth slice across the full scan depth). Predefined performance thresholds: mean DSC greater than or equal to 0.90 for skeletal muscle, subcutaneous adipose, and visceral adipose tissues; mean DSC greater than or equal to 0.85 for intramuscular adipose tissue. Thresholds must be met both in aggregate and within every demographic and clinical subgroup with at least 20 scans (BMI category, age band, sex, body region, clinical context). Reported with 95% bias-corrected and accelerated (BCa) bootstrap confidence intervals.
Time frame
Single time point: completion of standalone Soma inference and consolidated multi-rater annotation on all 200 study scans, anticipated within two weeks of study start.Secondary outcomes
1
Bland-Altman Agreement for Tissue Cross-Sectional Areas (cm^2)
Bland-Altman analysis comparing tissue cross-sectional areas (in cm\^2) computed by Soma versus the multi-rater radiologist reference, for each tissue class (skeletal muscle, subcutaneous adipose, visceral adipose, intramuscular adipose). Reported: mean bias and 95% limits of agreement, in aggregate and within demographic and clinical subgroups.
Time frame
Single time point: completion of Soma inference and consolidated multi-rater annotation on all 200 study scans, anticipated within two weeks of study start.2
95th-Percentile Hausdorff Distance Per Tissue Class
95th-percentile Hausdorff distance (HD95, in mm) between Soma segmentation contours and the multi-rater radiologist reference, computed per tissue class at the L3 vertebra. Reported in aggregate and within demographic and clinical subgroups.
Time frame
Single time point: completion of Soma inference and consolidated multi-rater annotation on all 200 study scans, anticipated within two weeks of study start.3
Pearson Correlation of Tissue Areas and Skeletal Muscle Index (SMI)
Pearson correlation coefficient (r) between Soma-derived and multi-rater-reference values for: (i) per-tissue cross-sectional areas (skeletal muscle, subcutaneous adipose, visceral adipose, intramuscular adipose, in cm\^2); and (ii) Skeletal Muscle Index (SMI = skeletal muscle area / height\^2, in cm\^2/m\^2). Reported with 95% confidence intervals, in aggregate and within subgroups.
Time frame
Single time point: completion of Soma inference and consolidated multi-rater annotation on all 200 study scans, anticipated within two weeks of study start.4
Cohen's Kappa for Sarcopenia Classification by Skeletal Muscle Index (SMI)
Cohen's kappa coefficient quantifying agreement between Soma-derived and reference-derived binary sarcopenia classification, defined by sex-specific Skeletal Muscle Index (SMI) cutoffs. Prespecified performance threshold: kappa greater than or equal to 0.80. Reported in aggregate and within demographic and clinical subgroups.
Time frame
Single time point: completion of Soma inference and consolidated multi-rater annotation on all 200 study scans, anticipated within two weeks of study start.Sponsors and contacts
Click on the lead sponsor to view all of their trials.
Source ClinicalTrials.gov
This trial is not recruiting at the moment. You can still explore other options: