Medical image processing applications typically demand highly accurate image segmentation. However, existing segmentation approaches exhibit performance degradation when faced with diverse medical imaging modalities and varied segmentation target sizes. In this paper, we propose and evaluate a dilated One-to-Many U-Net deep learning model that addresses these challenges. The proposed model comprises of four rows of encoder-decoder modules, with each module consisting of three trainable blocks with different layers. The last three rows of the U-Net are extended versions of the three blocks in the first row, with the encoder-decoder blocks connected through the skip connections to the previous rows. The outputs of the last blocks from the last three rows in the decoder are concatenated, and finally, a dilation network is employed to improve the small target segmentation in different medical images. Two datasets have been used for the evaluation: the HC18 grand challenge ultrasound dataset for fetal head segmentation and the Multi-site MRI dataset, including the BIDMC and HK sites, for prostate segmentation in MRI images. The proposed approach achieved Dice and Jaccard coefficients of 96.54% and 93.93%, respectively, for the HC18 grand challenge dataset, 96.76% and 93.97% for the BIDMC site dataset, and 92.58% and 86.96% for the HK site dataset. Statistical analyses showed that the proposed model outperformed several other U-Net-based models.

This study evaluated a deep-neural-network denoising system using model-based design, comparing it with adaptive filtering and beamforming across various noise types, SNRs, and hearing-aid fittings. A KEMAR manikin fitted with five audiograms was recorded in reverberant and non-reverberant rooms, yielding 1,152 recordings. Speech intelligibility was estimated using the HASPI from 1,152 KEMAR manikin recordings. Effects of processing strategy and acoustic factors were tested with model-based within-device design that account for repeated recordings per device/program and fitting. Linear mixed model results showed that the DNN with beamforming outperformed conventional processing, with strongest gains at 0 and +5 dB SNR, moderate benefits at −5 dB in low reverberation, and none in medium reverberation. Across SNRs and noise types, the DNN combined with beamforming yielded the highest predicted intelligibility, with benefits attenuated under moderate reverberation. Azimuth effects varied; because estimates were derived from a better-ear metric on manikin recordings. Additionally, this paper reports comparisons using metrics of sound quality, for an intrusive metric (HASQI) and the pMOS non-intrusive metric. Results indicated that model type interacted with processing and acoustic factors. HASQI and pMOS scores increased with SNR and were moderately correlated (r² ≈ 0.479), supporting the use of non-intrusive metrics for large-scale assessment. However, pMOS showed greater variability across hearing aid programs and environments, suggesting non-intrusive models capture processing effects differently than intrusive metrics. These findings highlight the promise and limits of non-intrusive evaluation while emphasizing the benefit of combining deep learning with beamforming to improve intelligibility and quality.

Recent advancements in speech enhancement have witnessed the emergence of generator-based methodologies. However, several of these approaches exhibit complexity in handling input variations, either excelling at low signal-to-noise ratios (SNRs) by utilizing intricate representations of noisy and clean speech or demonstrating superior performance only at higher SNRs. In this work, we investigated speech enhancement using a Dilated Attention Fast Generative Adversarial Network (DAF-GAN). The proposed DAF-GAN framework achieves stability in performance across different SNR conditions by efficiently processing large-scale signal lengths. The DFS-GAN features a dilated discriminator model operating via patches. The generator architecture incorporates multi-decoding and attention gates facilitated through skip-connections, strategically integrated within the Fast-U-Net model to optimize processing speed. An ideal ratio mask was used in the test phase to further refine the enhanced signal by emphasizing target speech while suppressing residual noise or artifacts. The DAF-GAN performance was assessed using objective metrics such as PESQ on a number of noisy speech databases. Results revealed that the DAF-GAN performed modestly in comparison with the state-of-the-art models. For example, analyses of the VoiceBank-DEMAND dataset yielded a PESQ score of 2.50 for the DAF-GAN.

In this study, we proposed a novel automated speech quality estimation model capable of evaluating perceptual dysphonia severity and breathiness in audio samples, ensuring alignment with expert-rated assessments. The proposed model integrates Whisper ASR embeddings with Mel spectrograms augmented by second-order delta features combined with a sequential-attention fusion network feature mapping path. This hybrid approach enhances the model’s sensitivity to phonetic, high level feature representation and spectral variations, enabling more accurate predictions of perceptual speech quality. A sequential-attention fusion network feature mapping module captures long-range de-pendencies through the multi-head attention network, while LSTM layers refine the learned representations by modeling temporal dynamics. Comparative analysis against state-of-the-art methods for dysphonia assessment demonstrates our model’s superior generalization across test samples. Our findings underscore the effectiveness of ASR-derived embeddings alongside the deep feature mapping structure in speech quality assessment, offering a promising pathway for advancing automated evaluation systems.

In this study, we propose a novel automated model for speech quality estimation that objectively evaluates perceptual dysphonia severity and breathiness in audio samples, demonstrating strong correlation with expert ratings. The proposed model integrates Whisper encoder embeddings with Mel spectrograms augmented by second-order delta features combined with a sequential-attention fusion network feature mapping path. This hybrid approach enhances the model’s sensitivity to phonetic, high-level feature representation, and spectral variations, enabling more accurate predictions of perceptual speech quality. A sequential-attention fusion network feature mapping module captures long-range dependencies through the multi-head attention network, while LSTM layers refine the learned representations by modeling temporal dynamics. Comparative analysis against state-of-the-art methods for dysphonia assessment demonstrates our model’s better correlation with clinician’s judgments across test samples. Our findings underscore the effectiveness of ASR-derived embeddings alongside the deep feature mapping structure in disordered speech quality assessment, offering a promising pathway for advancing automated evaluation systems.