A data-driven ultrasound approach discriminates pathological high grade prostate cancer

Jun Akatsuka,Hayato Takeda,Shigenori Kayama,Hikaru Mikami,Yoichiro Yamamoto,Yuka Toyama,Yukihiro Kondo,Hiromu Morikawa,Tetsuro Sekine,Yasushi Numata,Ruri Yamaguchi,Go Kimura,Yuki Endo,Masato Yanagi

doi:10.1038/s41598-022-04951-3

Jun Akatsuka, Hayato Takeda + Show 12 more

Open Access

https://doi.org/10.1038/s41598-022-04951-3

Copy DOI

Abstract

Accurate prostate cancer screening is imperative for reducing the risk of cancer death. Ultrasound imaging, although easy, tends to have low resolution and high inter-observer variability. Here, we show that our integrated machine learning approach enabled the detection of pathological high-grade cancer by the ultrasound procedure. Our study included 772 consecutive patients and 2899 prostate ultrasound images obtained at the Nippon Medical School Hospital. We applied machine learning analyses using ultrasound imaging data and clinical data to detect high-grade prostate cancer. The area under the curve (AUC) using clinical data was 0.691. On the other hand, the AUC when using clinical data and ultrasound imaging data was 0.835 (p = 0.007). Our data-driven ultrasound approach offers an efficient tool to triage patients with high-grade prostate cancers and expands the possibility of ultrasound imaging for the prostate cancer detection pathway.

Highlights

Accurate prostate cancer screening is imperative for reducing the risk of cancer death
Ultrasound imaging is widely used in prostate cancer screening because it is nonionizing, low-cost, and safe
We evaluated the prediction accuracies of the following different datasets: ultrasound image data, clinical data, and integrated data

Summary

Introduction

Accurate prostate cancer screening is imperative for reducing the risk of cancer death. We show that our integrated machine learning approach enabled the detection of pathological high-grade cancer by the ultrasound procedure. We applied machine learning analyses using ultrasound imaging data and clinical data to detect high-grade prostate cancer. Artificial intelligence (AI) technologies, including deep learning algorithms, are gaining extensive attention due to their excellent performance in image classification and object detection These algorithms have been useful tools in the analysis of medical images of various cancers, such as breast cancers[8], brain tumors[9], lung cancers[10], esophageal cancers[11], skin malignancies[12], and prostate cancers[13,14,15]. We aim to estimate pathological high-grade cancer using ordinary ultrasound images and limited clinical data

Objectives

Methods

Results

Conclusion