Deep Learning Assisted Localization of Polycystic Kidney on Contrast-Enhanced CT Images.

Djeane Debora Onthoni,Pushpanjali Gupta,Li-Jen Wang,Ting-Wen Sheng,Prasan Kumar Sahoo

doi:10.3390/diagnostics10121113

Djeane Debora Onthoni, Pushpanjali Gupta + Show 3 more

Open Access

https://doi.org/10.3390/diagnostics10121113

Copy DOI

Abstract

Total Kidney Volume (TKV) is essential for analyzing the progressive loss of renal function in Autosomal Dominant Polycystic Kidney Disease (ADPKD). Conventionally, to measure TKV from medical images, a radiologist needs to localize and segment the kidneys by defining and delineating the kidney’s boundary slice by slice. However, kidney localization is a time-consuming and challenging task considering the unstructured medical images from big data such as Contrast-enhanced Computed Tomography (CCT). This study aimed to design an automatic localization model of ADPKD using Artificial Intelligence. A robust detection model using CCT images, image preprocessing, and Single Shot Detector (SSD) Inception V2 Deep Learning (DL) model is designed here. The model is trained and evaluated with 110 CCT images that comprise 10,078 slices. The experimental results showed that our derived detection model outperformed other DL detectors in terms of Average Precision (AP) and mean Average Precision (mAP). We achieved mAP = 94% for image-wise testing and mAP = 82% for subject-wise testing, when threshold on Intersection over Union (IoU) = 0.5. This study proves that our derived automatic detection model can assist radiologist in locating and classifying the ADPKD kidneys precisely and rapidly in order to improve the segmentation task and TKV calculation.

Highlights

Autosomal Dominant Polycystic Kidney Disease (ADPKD) is the most common hereditary renal disease with an estimated prevalence of 1:1000 to 1:2500 [1,2]
We found that less work has been done to solve the detection problem in ADPKD kidneys using an object detection approach
It was observed that our model achieved the highest performance in various evaluation metrics with an accuracy of 0.90 for right and 0.91 for left kidney

Summary

Introduction

Autosomal Dominant Polycystic Kidney Disease (ADPKD) is the most common hereditary renal disease with an estimated prevalence of 1:1000 to 1:2500 [1,2]. Two important biomarkers need to be examined for predicting the progressive loss of renal function: Glomerular Filtration Rate (GFR) and Total Kidney Volume (TKV) [4]. TKV allows stratification of patients into low and high-risk subgroups to identify individuals who may benefit from the treatment [3,5]. TKV is calculated using common medical imaging tests such as Computed Tomography (CT) and Magnetic Resonance Imaging (MRI). Known for its faster and low-cost technique, CT has been used globally and is a popular technique in clinical imaging tests. Contrast-enhanced Computed Tomography (CCT) highlights the blood vessels and enhances organs which provides better contrast resolution on images as compared to Non-contrast-enhanced Computed Tomography (NCCT)

Objectives

Methods

Results

Discussion

Conclusion