Learn Fine-Grained Adaptive Loss for Multiple Anatomical Landmark Detection in Medical Images.

Automatic and accurate detection of anatomical landmarks is an essential operation in medical image analysis with a multitude of applications. Recent deep learning methods have improved results by directly encoding the appearance of the captured anatomy with the likelihood maps (i.e., heatmaps). However, most current solutions overlook another essence of heatmap regression, the objective metric for regressing target heatmaps and rely on hand-crafted heuristics to set the target precision, thus being usually cumbersome and task-specific. In this paper, we propose a novel learning-to-learn framework for landmark detection to optimize the neural network and the target precision simultaneously. The pivot of this work is to leverage the reinforcement learning (RL) framework to search objective metrics for regressing multiple heatmaps dynamically during the training process, thus avoiding setting problem-specific target precision. We also introduce an early-stop strategy for active termination of the RL agent's interaction that adapts the optimal precision for separate targets considering exploration-exploitation tradeoffs. This approach shows better stability in training and improved localization accuracy in inference. Extensive experimental results on two different applications of landmark localization: 1) our in-house prenatal ultrasound (US) dataset and 2) the publicly available dataset of cephalometric X-Ray landmark detection, demonstrate the effectiveness of our proposed method. Our proposed framework is general and shows the potential to improve the efficiency of anatomical landmark detection.

Effect of large-scale kelp and bivalve farming on seawater carbonate system variations in the semi-enclosed Sanggou Bay.

Although cultured algae and shellfish can be the dominant species in some localized coastal waters, research on the effect of large-scale mariculture on the carbonate system variations in these local waters is still lacking. We conducted five cruises from May to September and studied spatiotemporal variations in the seawater carbonate system in the semi-closed Sanggou Bay, which is famous for its large-scale mariculture. Our results showed that both kelp and bivalve farming induced significant spatiotemporal variations in the carbonate system within the bay. When cultured kelp reached its highest biomass in May, the maximum ΔDIC, ΔpCO2 and ΔpHT between the seawater from the kelp farming area and the non-farming outer bay area was -156 µmol kg-1, -102 µatm and 0.15 pH units, respectively. However, no significant effect of kelp farming on seawater total alkalinity (TA) was observed. Kelp farming also caused the carbonate system variations of seawater from the bivalve farming area. Assuming no kelp was farmed in May, the average pH and pCO2 would reduce by 0.12 pH units and increase by 179 µatm, respectively, in the bivalve farming area. Bivalve farming significantly reduced seawater TA, indicating that fast deposition of calcium carbonate occurred in the bivalve farming area. Although bivalve respiration released CO2 into seawater and elevated seawater pCO2 level and reduced seawater pHT, surprisingly, seawater dissolved inorganic carbon (DIC) reduced significantly in the bivalve farming area. These results indicated that bivalves fixed a larger amount of inorganic carbon by calcification than that released into seawater by respiration. Overall, large-scale kelp and bivalve farming are important biological drivers of variations in the carbonate system within the semi-enclosed Sanggou Bay. Altered carbonate systems by kelp farming may favour calcification of farmed bivalves and provide an essential refuge for these species during the future ocean acidification.

A Single-Shot Region-Adaptive Network for Myotendinous Junction Segmentation in Muscular Ultrasound Images.

Tracking the myotendinous junction (MTJ) in consecutive ultrasound images is crucial for understanding the mechanics and pathological conditions of the muscle-tendon unit. However, the lack of reliable and efficient identification of MTJ due to poor image quality and boundary ambiguity restricts its application in motion analysis. In recent years, with the rapid development of deep learning, the region-based convolution neural network (RCNN) has shown great potential in the field of simultaneous objection detection and instance segmentation in medical images. This article proposes a region-adaptive network (RAN) to localize MTJ region and to segment it in a single shot. Our model learns about the salient information of MTJ with the help of a composite architecture. Herein, a region-based multitask learning network explores the region containing MTJ, while a parallel end-to-end U-shaped path extracts the MTJ structure from the adaptively selected region for combating data imbalance and boundary ambiguity. By demonstrating the ultrasound images of the gastrocnemius, we showed that the RAN achieves superior segmentation performance when compared with the state-of-the-art Mask RCNN method with an average Dice score of 80.1%. Our proposed method is robust and reliable for advanced muscle and tendon function examinations obtained by ultrasound imaging.