Dataset Information

Deep Learning Automation of Kidney, Liver, and Spleen Segmentation for Organ Volume Measurements in Autosomal Dominant Polycystic Kidney Disease.

ABSTRACT: Organ volume measurements are a key metric for managing ADPKD (the most common inherited renal disease). However, measuring organ volumes is tedious and involves manually contouring organ outlines on multiple cross-sectional MRI or CT images. The automation of kidney contouring using deep learning has been proposed, as it has small errors compared to manual contouring. Here, a deployed open-source deep learning ADPKD kidney segmentation pipeline is extended to also measure liver and spleen volumes, which are also important. This 2D U-net deep learning approach was developed with radiologist labeled T2-weighted images from 215 ADPKD subjects (70% training = 151, 30% validation = 64). Additional ADPKD subjects were utilized for prospective (n = 30) and external (n = 30) validations for a total of 275 subjects. Image cropping previously optimized for kidneys was included in training but removed for the validation and inference to accommodate the liver which is closer to the image border. An effective algorithm was developed to adjudicate overlap voxels that are labeled as more than one organ. Left kidney, right kidney, liver and spleen labels had average errors of 3%, 7%, 3%, and 1%, respectively, on external validation and 5%, 6%, 5%, and 1% on prospective validation. Dice scores also showed that the deep learning model was close to the radiologist contouring, measuring 0.98, 0.96, 0.97 and 0.96 on external validation and 0.96, 0.96, 0.96 and 0.95 on prospective validation for left kidney, right kidney, liver and spleen, respectively. The time required for manual correction of deep learning segmentation errors was only 19:17 min compared to 33:04 min for manual segmentations, a 42% time saving (p = 0.004). Standard deviation of model assisted segmentations was reduced to 7, 5, 11, 5 mL for right kidney, left kidney, liver and spleen respectively from 14, 10, 55 and 14 mL for manual segmentations. Thus, deep learning reduces the radiologist time required to perform multiorgan segmentations in ADPKD and reduces measurement variability.

SUBMITTER: Sharbatdaran A

PROVIDER: S-EPMC9326744 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundIn autosomal dominant polycystic kidney disease (ADPKD), total kidney volume (TKV) is regarded as an important biomarker of disease progression and different methods are available to assess kidney volume. The purpose of this study was to identify the most efficient kidney volume computation method to be used in clinical studies evaluating the effectiveness of treatments on ADPKD progression.Methods and findingsWe measured single kidney volume (SKV) on two series of MR and CT images from clinical studies on ADPKD (experimental dataset) by two independent operators (expert and beginner), twice, using all of the available methods: polyline manual tracing (reference method), free-hand manual tracing, semi-automatic tracing, Stereology, Mid-slice and Ellipsoid method. Additionally, the expert operator also measured the kidney length. We compared different methods for reproducibility, accuracy, precision, and time required. In addition, we performed a validation study to evaluate the sensitivity of these methods to detect the between-treatment group difference in TKV change over one year, using MR images from a previous clinical study. Reproducibility was higher on CT than MR for all methods, being highest for manual and semiautomatic contouring methods (planimetry). On MR, planimetry showed highest accuracy and precision, while on CT accuracy and precision of both planimetry and Stereology methods were comparable. Mid-slice and Ellipsoid method, as well as kidney length were fast but provided only a rough estimate of kidney volume. The results of the validation study indicated that planimetry and Stereology allow using an importantly lower number of patients to detect changes in kidney volume induced by drug treatment as compared to other methods.ConclusionsPlanimetry should be preferred over fast and simplified methods for accurately monitoring ADPKD progression and assessing drug treatment effects. Expert operators, especially on MR images, are required for performing reliable estimation of kidney volume. The use of efficient TKV quantification methods considerably reduces the number of patients to enrol in clinical investigations, making them more feasible and significant.

Project description:Background and objectivesTo evaluate the growth pattern of kidney cyst number and cyst volume in association with kidney size, demographics, and genotypes in autosomal dominant polycystic kidney disease.Design, setting, participants, & measurementsKidney cyst number and cyst volume were measured from serial magnetic resonance images, giving a maximum follow-up of 14.23 years, from 241 patients with autosomal dominant polycystic kidney disease (15-46 years old at baseline). The growth pattern was analyzed, in association with sex, age, height-adjusted total kidney volume, and genotype, using linear mixed models of repeated measurements and tests of interactions with age (as a time-dependent covariate) to assess rates of change over time. Models were also fit using Irazabal class. Genotypic groups were characterized as either (1) PKD1 truncating, PKD1 nontruncating, and PKD2 plus patients with no mutation detected; or (2) in combination with PKD1 mutation strength groups.ResultsImaging and genetic data were collected (at least one visit) for 236 participants. The mean height-adjusted total cyst number increased exponentially over time from a baseline value of 762 to 1715 at the last clinic visit, while the mean height-adjusted total cyst volume increased exponentially from 305 to 770 ml. Height-adjusted total kidney volume, height-adjusted total cyst number, and height-adjusted total cyst volume were all highly correlated over time. Female participants and participants with larger height-adjusted total kidney volume at baseline showed smaller rates of change in the log of height-adjusted total cyst number and cyst volume. PKD1 was associated with significant increases in both cyst number and volume at a given age, but genotype did not significantly affect the rate of growth.ConclusionsBoth height-adjusted total cyst number and height-adjusted total cyst volume increased exponentially and more than doubled over 14.23 years of follow-up. Compared with PKD2 plus no mutation detected, PKD1 was associated with a greater cyst number and volume at a given age, but no significant difference in the rate of growth.

Project description:Background and objectivesThe progression of polycystic liver disease is not well understood. The purpose of the study is to evaluate the associations of polycystic liver progression with other disease progression variables and classify liver progression on the basis of patient's age, height-adjusted liver cystic volume, and height-adjusted liver volume.Design, setting, participants, & measurementsProspective longitudinal magnetic resonance images from 670 patients with early autosomal dominant polycystic kidney disease for up to 14 years of follow-up were evaluated to measure height-adjusted liver cystic volume and height-adjusted liver volume. Among them, 245 patients with liver cyst volume >50 ml at baseline were included in the longitudinal analysis. Linear mixed models on log-transformed height-adjusted liver cystic volume and height-adjusted liver volume were fitted to approximate mean annual rate of change for each outcome. The association of sex, body mass index, genotype, baseline height-adjusted total kidney volume, and Mayo imaging class was assessed. We calculated height-adjusted liver cystic volume ranges for each specific age and divided them into five classes on the basis of annual percentage increase in height-adjusted liver cystic volume.ResultsThe mean annual growth rate of height-adjusted liver cystic volume was 12% (95% confidence interval, 11.1% to 13.1%; P<0.001), whereas that for height-adjusted liver volume was 2% (95% confidence interval, 1.9% to 2.6%; P<0.001). Women had higher baseline height-adjusted liver cystic volume than men, but men had higher height-adjusted liver cystic volume growth rate than women by 2% (95% confidence interval, 0.4% to 4.5%; P=0.02). Whereas the height-adjusted liver cystic volume growth rate decreased in women after menopause, no decrease was observed in men at any age. Body mass index, genotype, and baseline height-adjusted total kidney volume were not associated with the growth rate of height-adjusted liver cystic volume or height-adjusted liver volume. According to the height-adjusted liver cystic volume growth rate, patients were classified into five classes (number of women, men in each class): A (24, six); B (44, 13); C (43, 48); D (28, 17); and E (13, nine).ConclusionsCompared with height-adjusted liver volume, the use of height-adjusted liver cystic volume showed greater separations in volumetric progression of polycystic liver disease. Similar to the Mayo imaging classification for the kidney, the progression of polycystic liver disease may be categorized on the basis of patient's age and height-adjusted liver cystic volume.

Project description:Background and objectivesOur study developed a fully automated method for segmentation and volumetric measurements of kidneys from magnetic resonance images in patients with autosomal dominant polycystic kidney disease and assessed the performance of the automated method with the reference manual segmentation method.Design, setting, participants, & measurementsStudy patients were selected from the Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease. At the enrollment of the Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease Study in 2000, patients with autosomal dominant polycystic kidney disease were between 15 and 46 years of age with relatively preserved GFRs. Our fully automated segmentation method was on the basis of a spatial prior probability map of the location of kidneys in abdominal magnetic resonance images and regional mapping with total variation regularization and propagated shape constraints that were formulated into a level set framework. T2-weighted magnetic resonance image sets of 120 kidneys were selected from 60 patients with autosomal dominant polycystic kidney disease and divided into the training and test datasets. The performance of the automated method in reference to the manual method was assessed by means of two metrics: Dice similarity coefficient and intraclass correlation coefficient of segmented kidney volume. The training and test sets were swapped for crossvalidation and reanalyzed.ResultsSuccessful segmentation of kidneys was performed with the automated method in all test patients. The segmented kidney volumes ranged from 177.2 to 2634 ml (mean, 885.4±569.7 ml). The mean Dice similarity coefficient ±SD between the automated and manual methods was 0.88±0.08. The mean correlation coefficient between the two segmentation methods for the segmented volume measurements was 0.97 (P<0.001 for each crossvalidation set). The results from the crossvalidation sets were highly comparable.ConclusionsWe have developed a fully automated method for segmentation of kidneys from abdominal magnetic resonance images in patients with autosomal dominant polycystic kidney disease with varying kidney volumes. The performance of the automated method was in good agreement with that of manual method.

Dataset Information

Deep Learning Automation of Kidney, Liver, and Spleen Segmentation for Organ Volume Measurements in Autosomal Dominant Polycystic Kidney Disease.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets