Decision support with MRI for targeting, evaluating laser ablation for prostate c

描述（由申请人提供）：对于不断增长的低风险和中风险前列腺癌（CaP）患者人群，主动监测可能在生物学或心理学上不受欢迎。然而，与根治性全器官治疗（例如调强放射治疗）相关的短期和长期并发症和合并症仍与治疗相关发病风险相关。在某些患者中， 对放射治疗较敏感的病人，直肠和膀胱内的高剂量点可导致并发症，包括慢性直肠出血，腹泻和膀胱炎等泌尿系统症状。激光诱导间质热疗（LITT）是一种新型的受控靶向热消融，与其他消融疗法相比，可为前列腺局灶性治疗提供可测量的优势。由于LITT与MRI兼容，因此与其他利用经直肠超声定位和监测治疗的手术或消融技术相比，它具有成像优势。然而，成功采用局灶性疗法治疗CaP将取决于几个关键问题：1）我们能否准确识别前列腺内的索引病变和癌症？2)对接受局部治疗的患者进行适当的随访; 3）如何检测复发/持续性疾病？多参数（MP）MRI（T2 w、动态对比增强（DCE）、弥散（DWI））的引入允许（1）提高CaP定位的检测灵敏度和特异性，以及（2）评价前列腺中的治疗反应。然而，需要（1）新的计算图像分析工具来定量整合MP-MRI参数，以改善体内CaP分类，以及（2）非刚性配准工具，以实现靶向治疗和治疗后变化的评估。 在本研究中，我们将采用复杂的计算机视觉、图像分析、计算机辅助诊断（CAD）和可变形配准工具与MP-MRI结合使用，与涉及40名患有记录的CaP患者的小型临床试验结合使用，用于（a）自动描绘肿瘤区域治疗前MP-MRI，从而通过LITT识别消融的特定区域，和（B）识别和描绘消融区内外的局部复发性疾病，用于LITT后评估。在LITT前MRI上通过CAD识别的区域将成为治疗的目标，而在LITT后MP-MRI上，将通过针芯活检评价被识别为疑似CaP复发的区域（由于MR成像标记物的大幅变化）。本项目中开发的工具将被整合到一个实际可行的治疗范例中，用于低风险局部CaP的局部治疗，这将使患者避免与根治性全腺体治疗相关的并发症。这个跨学科的翻译项目结合了MP-MRI CAD、多模式图像配准和机器学习方面的工程专业知识，以及介入放射学、前列腺MRI和MRI引导的病灶治疗方面的临床专业知识。 公共卫生相关性：在这项研究中，我们将利用先进的计算机视觉、图像分析、计算机辅助诊断和可变形配准工具，结合前列腺癌（CaP）患者的多参数（MP）MRI，用于（a）在治疗前MP-MRI上自动描绘肿瘤区域，从而识别通过激光诱导间质热治疗（LITT）进行消融的特定区域，和（B）识别和描绘消融区内外的局部复发性疾病，用于LITT后评估。本项目中开发的工具将被整合到一个可行的治疗范例中用于病灶治疗 低风险的局部CaP，这将使患者避免与根治性全腺体治疗相关的并发症。

DESCRIPTION (provided by applicant): For a growing population of low- and intermediate-risk prostate cancer (CaP) patients, active surveillance may be biologically or psychologically undesirable. Yet the short- and long-term complications and co-morbidities associated with radical whole-organ therapies (e.g. intensity modulated radiation therapy) are still associated with a risk of treatment-related morbidity. With radical whole gland therapies, in certain patients who are more sensitive to radiation treatment, high dosage spots in the rectum and in the bladder can lead to complications including chronic rectal bleeding, diarrhea, and urinary symptoms such as cystitis. Laser induced interstitial thermal therapy (LITT) is a novel form of controlled, targeted thermal ablation that may offer measurable advantages over other ablative therapies for focal prostate therapy. Because LITT is MRI compatible, it enables an imaging advantage over other surgical or ablation techniques that utilize transrectal ultrasound to target and monitor treatment. However successful adoption of focal therapy for the treatment of CaP will hinge on several critical issues: 1) Can we accurately identify index lesions and cancers within the prostate? 2) Appropriate follow-up of patients treated with focal therapy and 3) How to detect recurrent/persistent disease? The introduction of multi-parametric (MP) MRI (T2w, dynamic contrast enhanced (DCE), Diffusion (DWI)) has allowed for (1) improved detection sensitivity and specificity for CaP localization, and (2) evaluating treatment response in the prostate. However there exists a need for (1) novel computational image analysis tools to quantitatively integrate MP-MRI parameters for improved CaP classification in vivo and (2) non-rigid registration tools for enabling targeted therapy and evaluation of post-treatment changes. In this study we will employ sophisticated computer vision, image analysis, computer assisted diagnostic (CAD) and deformable registration tools in conjunction with MP-MRI to be used in conjunction with a small clinical trial involving 40 patients with documented CaP for (a) automated delineation of tumor regions on pre-treatment MP-MRI to thereby identify the specific regions for ablation via LITT, and (b) identify and delineate locally recurrent disease within and outside the ablation zone for post-LITT evaluation. Regions identified via CAD on pre-LITT MRI will be targeted for therapy, while on post-LITT MP-MRI, regions identified as being suspicious for CaP recurrence (on account of large changes in MR imaging markers) will be evaluated via needle core biopsy. The tools developed in this project will be integrated into a practical and feasible treatment paradigm for focal treatment of low-risk localized CaP which will allow patients to avoid the complications associated with radical whole-gland therapy. This inter-disciplinary, translational project combines engineering expertise in terms of CAD on MP-MRI, multimodal image registration and machine learning and clinical expertise in interventional radiology, prostate MRI, and MRI guided focal therapy. PUBLIC HEALTH RELEVANCE: In this study we will leverage sophisticated computer vision, image analysis, computer assisted diagnostic and deformable registration tools in conjunction with multi-parametric (MP) MRI in prostate cancer (CaP) patients for (a) automated delineation of tumor regions on pre-treatment MP-MRI and thereby identify the specific regions for ablation via laser induced interstitial thermal therapy (LITT), and (b) identify and delineate locally recurrent disease within and outside the ablation zone for post-LITT evaluation. The tools developed in this project will be integrated into a feasible treatment paradigm for focal treatment of low-risk localized CaP which will allow patients to avoid the complications associated with radical whole-gland therapy.