Brain image segmentation algorithm based on deep learning is a research hotspot at present. In this paper, firstly, the significance of brain image segmentation and the content of related brain image segmentation algorithm are systematically described, highlighting the advantages of brain image segmentation algorithms based on deep learning. Then, this paper introduces current brain image segmentation algorithms based on deep learning from three aspects: the brain image segmentation algorithms based on problems existent to brain image, the brain image segmentation algorithms based on prior knowledge guidance and the application of general deep learning models in brain image segmentation, so as to enable researchers in relevant fields to understand current research progress more systematically. Finally, this paper provides a general direction for the further research of brain image segmentation algorithm based on deep learning.
ObjectiveTo evaluate the clinical efficacy and safety of directional atherectomy (DA) combined with drug-coated balloon (DCB) for the treatment of femoropopliteal artery lesions. MethodsA retrospective analysis was performed on data from patients with femoropopliteal artery lesions treated with DA combined with paclitaxel DCB from January 2021 to November 2022 in the TALENT study (NCT04675632). Observational metrics included intraoperative bailout stenting rate, maximum DCB diameter, 12-month freedom from clinically-driven target limb revascularization (CD-TLR), incidence of major adverse events (MAEs), and changes in Rutherford category and vascular-related quality of life score. ResultsThis study enrolled 189 limbs from 184 patients. The age was (71.0±9.4) years, and 69.0% (127/184) were male. Comorbidities were common, including diabetes mellitus [71.2% (131/184)] and coronary artery disease [34.2% ( 63/184)], with 32.6% (60/184) having a history of smoking. The lesions had a length of (181.2±119.6) mm, with chronic total occlusion accounting for 63.0% (119/189). In terms of clinical severity, 47.1% (89/189) were classified as Rutherford category 3, and 52.9% (100/189) presented with chronic limb-threatening ischemia (Rutherford categories 4–6). According to Trans-Atlantic Inter-Society Consensus Ⅱ classification, 52.9% (100/189) were type A+B, and 47.1% (89/189) were type C+D; 18.0%(34/189) of the lesions exhibited severe calcification. The technical success rate of DA combined with DCB treatment was 96.8%(183/189), with a bailout stent implantation rate of 4.2% (8/189). The maximum DCB diameter used was (5.1±0.6) mm. At the 12-month follow-up, the cumulative freedom from CD-TLR rate was 96.1% [95%CI (92.6%, 98.2%)], and the cumulative freedom from MAEs rate was 89.2% [95%CI (84.0%, 92.9%)]. Both the postoperative Rutherford category and quality of life score showed significant improvement compared to baseline [1.81±1.39 versus 3.89±1.22, MD (95%CI)=?2.08 (?2.29, ?1.87), P<0.001; (5.51±1.09) points versus (2.90±1.05) points, MD (95%CI)=2.61 (2.46, 2.76), P<0.001]. ConclusionDA combined with paclitaxel DCB for the treatment of femoropopliteal artery lesions effectively achieves lumen enlargement, reduces the need for intraoperative bailout stenting, and demonstrates favorable 12-month clinical outcomes, representing a feasible vessel preparation strategy.
