Segmentation and three-dimensional reconstruction of lesions using the automated breast volume scanner (ABVS)

Dilated densely connected U-Net with uncertainty focus loss for 3D ABUS mass segmentation

BACKGROUND AND OBJECTIVE

Accurate segmentation of breast mass in 3D automated breast ultrasound (ABUS) images plays an important role in qualitative and quantitative ABUS image analysis. Yet this task is challenging due to the low signal to noise ratio and serious artifacts in ABUS images, the large shape and size variation of breast masses, as well as the small training dataset compared with natural images. The purpose of this study is to address these difficulties by designing a dilated densely connected U-Net (D²U-Net) together with an uncertainty focus loss.

METHODS

A lightweight yet effective densely connected segmentation network is constructed to extensively explore feature representations in the small ABUS dataset. In order to deal with the high variation in shape and size of breast masses, a set of hybrid dilated convolutions is integrated into the dense blocks of the D²U-Net. We further suggest an uncertainty focus loss to put more attention on unreliable network predictions, especially the ambiguous mass boundaries caused by low signal to noise ratio and artifacts. Our segmentation algorithm is evaluated on an ABUS dataset of 170 volumes from 107 patients. Ablation analysis and comparison with existing methods are conduct to verify the effectiveness of the proposed method.

RESULTS

Experiment results demonstrate that the proposed algorithm outperforms existing methods on 3D ABUS mass segmentation tasks, with Dice similarity coefficient, Jaccard index and 95% Hausdorff distance of 69.02%, 56.61% and 4.92 mm, respectively.

CONCLUSIONS

The proposed method is effective in segmenting breast masses on our small ABUS dataset, especially breast masses with large shape and size variations.

Uncertainty Aware Temporal-Ensembling Model for Semi-Supervised ABUS Mass Segmentation

Accurate breast mass segmentation of automated breast ultrasound (ABUS) images plays a crucial role in 3D breast reconstruction which can assist radiologists in surgery planning. Although the convolutional neural network has great potential for breast mass segmentation due to the remarkable progress of deep learning, the lack of annotated data limits the performance of deep CNNs. In this article, we present an uncertainty aware temporal ensembling (UATE) model for semi-supervised ABUS mass segmentation. Specifically, a temporal ensembling segmentation (TEs) model is designed to segment breast mass using a few labeled images and a large number of unlabeled images. Considering the network output contains correct predictions and unreliable predictions, equally treating each prediction in pseudo label update and loss calculation may degrade the network performance. To alleviate this problem, the uncertainty map is estimated for each image. Then an adaptive ensembling momentum map and an uncertainty aware unsupervised loss are designed and integrated with TEs model. The effectiveness of the proposed UATE model is mainly verified on an ABUS dataset of 107 patients with 170 volumes, including 13382 2D labeled slices. The Jaccard index (JI), Dice similarity coefficient (DSC), pixel-wise accuracy (AC) and Hausdorff distance (HD) of the proposed method on testing set are 63.65%, 74.25%, 99.21% and 3.81mm respectively. Experimental results demonstrate that our semi-supervised method outperforms the fully supervised method, and get a promising result compared with existing semi-supervised methods.

Breast cancer research and treatment reconstruction of unilateral breast structure using three-dimensional ultrasound imaging to assess breast neoplasm

Purpose

To develop and evaluate the accuracy of a three-dimensional (3D) US method for assessing unilateral breast reconstruction and discuss the feasibility of breast ultrasound 3D reconstruction of the unilateral breast compared with 3D MRI.

Methods

Sixty-four breast lesions were collected for surgical resection. (1) MRI and US imaging were used to reconstruct the 3D models of the breast neoplasm. The diameters for maximum length, width, and depth of the negative tumor margins were used as the primary standards for comparison. (2) The measurement direction was determined by the largest gravity change between the two body positions. (3) The vertical distance from the midpoint of breast neoplasm to the ipsilateral nipple was calculated via MRI and US reconstruction.

Results

(1) Comparison of the measured size and histopathology of the breast neoplasm showed that US, MRI, and histopathology were highly correlated (p < 0.001). (2) When compared with the other two vertical directions, the direction with the largest gravity change had the greatest difference between MRI and US measurements. (3) The vertical distance from the breast neoplasm to the ipsilateral nipple and skin junction was significantly different (p > 0.05).

Conclusions

We have presented a novel US 3D reconstruction method for evaluating tumor size, which can provide a basis for investigated advanced visualization techniques for assessing breast tissue such as holographic presentation of 3D image data. These methods can provide physicians with a novel approach for making accurate surgical plans, for better communication with patients, and for more effective navigating throughout the operation.

A novel planning solution for semi-autonomous aspiration of Baker's cysts

BACKGROUND

A Baker's cyst is a pathological structure located near a kneepit, which causes discomfort and reduces mobility of the knee. It is commonly treated with aspiration, which often requires MRI scanning and US guidance. The aim of this study was to propose a novel planning solution for semi-autonomous aspiration of the Baker's cyst using only MRI imaging.

METHODS

The proposed method requires minimal user input and offers automatic cyst segmentation with collision-free path planning for the assumed robotic structure with four degrees of freedom.

RESULTS

The prepared software was tested on four image sets obtained from patients eligible for cyst aspiration. It was possible to accurately segment the cyst in the considered cases. The collision-free path planning method was investigated in numerical scenarios.

CONCLUSIONS

The simulations verified the proposed software solution. Future work will be devoted to experimental verification of the path planning procedure.