Medical Physics

New submissions

• New submissions
• Cross-lists

New submissions for Wed, 17 Aug 22

[1]  arXiv:2208.07718 [pdf, ps, other]

Title: The depth of the banana and the impulse stripe illumination for diffuse optical tomography

Authors: Manabu Machida, Keita Osada, Keiichiro Kagawa

Subjects: Medical Physics (physics.med-ph); Optics (physics.optics)

The stripe illumination lies between the illumination in the spatial-frequency domain and the point illumination. Although the stripe illumination has a periodic structure as the illumination in the spatial-frequency domain, light from the stripe illumination can reach deep regions in biological tissue since it can be regarded as an array of point illuminations. For a pair of a source and a detector, the shape of light paths which connect the source and detector is called the banana shape. First, we investigate the depth of the banana. In the case of the zero boundary condition, we found that the depth of the center of the banana is $d_{\rm SD}/(2\sqrt{2})$, where $d_{\rm SD}$ is the distance between the source and detector on the boundary. In general, the depth is about $0.3d_{\rm SD}$ when $d_{\rm SD}\approx2\,{\rm cm}$ and the ratio of refractive indices on the boundary is about $1.37$. Next, we perform diffuse optical tomography for the stripe illumination against forward data taken by Monte Carlo simulation. We consider an impulse illumination of the shape of a stripe. This time-resolved measurement is more informative than conventional continuous-wave measurements in the spatial-frequency domain.

Cross-lists for Wed, 17 Aug 22

[2]  arXiv:2208.07655 (cross-list from eess.IV) [pdf, other]

Title: A Hybrid Deep Feature-Based Deformable Image Registration Method for Pathological Images

Authors: Chulong Zhang, Yuming Jiang, Na Li, Zhicheng Zhang, Md Tauhidul Islam, Jingjing Dai, Lin Liu, Wenfeng He, Wenjian Qin, Jing Xiong, Yaoqin Xie, Xiaokun Liang

Comments: 23 pages, 12 figures

Subjects: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Medical Physics (physics.med-ph)

Pathologists need to combine information from differently stained pathological slices to obtain accurate diagnostic results. Deformable image registration is a necessary technique for fusing multi-modal pathological slices. This paper proposes a hybrid deep feature-based deformable image registration framework for stained pathological samples. We first extract dense feature points and perform points matching by two deep learning feature networks. Then, to further reduce false matches, an outlier detection method combining the isolation forest statistical model and the local affine correction model is proposed. Finally, the interpolation method generates the DVF for pathology image registration based on the above matching points. We evaluate our method on the dataset of the Non-rigid Histology Image Registration (ANHIR) challenge, which is co-organized with the IEEE ISBI 2019 conference. Our technique outperforms the traditional approaches by 17% with the Average-Average registration target error (rTRE) reaching 0.0034. The proposed method achieved state-of-the-art performance and ranking it 1 in evaluating the test dataset. The proposed hybrid deep feature-based registration method can potentially become a reliable method for pathology image registration.

• New submissions
• Cross-lists