KM Research Hub

P1 · PANEL / ROI / AXIS GROUND TRUTH

原 EXE 对齐基准真值复核

所有保存均写入追加式审计事件;旧标注、旧比较和旧提取结果不会被删除或覆盖。

返回质量诊断
冻结记录120
真值完整117
待复核0
无效/退役3
← 上一例 Case 64 / 120 · DB #64 下一例 →
multi_panel pending PMID 42313516

A Randomized Controlled Phase 3 Trial of Oral Prednisolone Administration Versus Local Triamcinolone Injection Therapy for Esophageal Stricture Prevention After Extensive Endoscopic Submucosal Dissection (JCOG1217).

den70195-fig-0003 · Panel B · Stricture-free survival

(a) Stricture‐free survival (SFS) for lesions with luminal circumference of ≥ 3/4 and < 7/8. The 12‐week SFS of the cases with mucosal defect over three‐fourths to seven‐eighths of the lumen was 96.0% (95% CI: 84.9–99.0) and 96.5% (95% CI: 86.7–99.1) in the local triamcinolone arm and the oral prednisolone arm, respectively (HR 1.229, 95% CI [0.275–5.492]). (b) SFS for the lesions with luminal circumference of > 7/8 but not the whole circumference. The SFS at 12 weeks was 75.2% (95% CI: 58.8–85.8) and 92.5% (95% CI: 78.5–97.5) in the local triamcinolone arm and the oral prednisolone arm, respectively (HR 0.545, 95% CI [0.198–1.502]). (c) SFS for the lesions with a whole‐circumference mucosal defect. The SFS at 12 weeks was 50.0% (95% CI: 11.1–80.4) and 75.0% (95% CI: 12.8–96.1) in the local triamcinolone arm and the oral prednisolone arm, respectively (HR 0.360, 95% CI [0.040–3.223]).
打开论文记录

坐标框复核

洋红 = 当前 Linux 候选;绿色 = 已保存人工真值。没有绿色框表示仍未形成真值。

Case 64 ROI review overlay
当前候选 {'x0': 74, 'y0': 11, 'x1': 697, 'y1': 336} 人工真值 {'x0': 74, 'x1': 697, 'y0': 11, 'y1': 336}
生产预期曲线2
当前恢复曲线2
算法版本legacy-0.5.0
坐标 OCR不完整/待核
查看候选 OCR 与警告(只能作种子,不能自证真值)
{
  "complete": false,
  "engine": "RapidOCR",
  "plot_bottom": 328,
  "plot_left": 88,
  "plot_right": 686,
  "plot_top": 90,
  "warnings": [
    "insufficient_x_ticks"
  ],
  "x_axis_y": 328,
  "x_confidence": 0.0,
  "x_detected": false,
  "x_max": null,
  "x_min": null,
  "x_ticks": [],
  "y_axis_x": 88,
  "y_confidence": 0.998,
  "y_detected": true,
  "y_max": 1,
  "y_min": 0,
  "y_ticks": [
    {
      "box": [
        [
          55.0,
          81.0
        ],
        [
          76.0,
          81.0
        ],
        [
          76.0,
          98.0
        ],
        [
          55.0,
          98.0
        ]
      ],
      "confidence": 0.9984111587206522,
      "pixel": 89.5,
      "text": "1.0",
      "value": 1.0
    },
    {
      "box": [
        [
          55.0,
          106.0
        ],
        [
          76.0,
          106.0
        ],
        [
          76.0,
          122.0
        ],
        [
          55.0,
          122.0
        ]
      ],
      "confidence": 0.9989785154660543,
      "pixel": 114.0,
      "text": "0.9",
      "value": 0.9
    },
    {
      "box": [
        [
          54.0,
          129.0
        ],
        [
          76.0,
          129.0
        ],
        [
          76.0,
          146.0
        ],
        [
          54.0,
          146.0
        ]
      ],
      "confidence": 0.998770554860433,
      "pixel": 137.5,
      "text": "0.8",
      "value": 0.8
    },
    {
      "box": [
        [
          55.0,
          153.0
        ],
        [
          76.0,
          153.0
        ],
        [
          76.0,
          169.0
        ],
        [
          55.0,
          169.0
        ]
      ],
      "confidence": 0.999228835105896,
      "pixel": 161.0,
      "text": "0.7",
      "value": 0.7
    },
    {
      "box": [
        [
          54.0,
          178.0
        ],
        [
          76.0,
          178.0
        ],
        [
          76.0,
          194.0
        ],
        [
          54.0,
          194.0
        ]
      ],
      "confidence": 0.9982431133588155,
      "pixel": 186.0,
      "text": "0.6",
      "value": 0.6
    },
    {
      "box": [
        [
          55.0,
          200.0
        ],
        [
          76.0,
          200.0
        ],
        [
          76.0,
          217.0
        ],
        [
          55.0,
          217.0
        ]
      ],
      "confidence": 0.9989339709281921,
      "pixel": 208.5,
      "text": "0.5",
      "value": 0.5
    },
    {
      "box": [
        [
          55.0,
          224.0
        ],
        [
          76.0,
          224.0
        ],
        [
          76.0,
          241.0
        ],
        [
          55.0,
          241.0
        ]
      ],
      "confidence": 0.9988491535186768,
      "pixel": 232.5,
      "text": "0.4",
      "value": 0.4
    },
    {
      "box": [
        [
          54.0,
          248.0
        ],
        [
          76.0,
          248.0
        ],
        [
          76.0,
          265.0
        ],
        [
          54.0,
          265.0
        ]
      ],
      "confidence": 0.9972966909408569,
      "pixel": 256.5,
      "text": "0.3",
      "value": 0.3
    },
    {
      "box": [
        [
          55.0,
          272.0
        ],
        [
          76.0,
          272.0
        ],
        [
          76.0,
          289.0
        ],
        [
          55.0,
          289.0
        ]
      ],
      "confidence": 0.998908797899882,
      "pixel": 280.5,
      "text": "0.2",
      "value": 0.2
    },
    {
      "box": [
        [
          54.0,
          297.0
        ],
        [
          76.0,
          297.0
        ],
        [
          76.0,
          313.0
        ],
        [
          54.0,
          313.0
        ]
      ],
      "confidence": 0.9978393912315369,
      "pixel": 305.0,
      "text": "0.1",
      "value": 0.1
    },
    {
      "box": [
        [
          54.0,
          320.0
        ],
        [
          76.0,
          320.0
        ],
        [
          76.0,
          336.0
        ],
        [
          54.0,
          336.0
        ]
      ],
      "confidence": 0.9962128003438314,
      "pixel": 328.0,
      "text": "0.0",
      "value": 0.0
    }
  ]
}
查看当前 Linux 输出

追加一条审查结论

选择“有效 KM”时,四边 ROI、四个坐标范围、Panel、终点和曲线数均为必填。

像素 ROI(x1/y1 为排他边界)
真实坐标范围

追加式审计历史

最新事件仅作为当前缓存;以下每一条既往结论都永久保留。

事件时间 / 审查人状态ROI坐标说明
#67p1-visual-axis-audit-v2 · frozen benchmark contact sheets + original images + RapidOCR re-audit 2026-07-28T15:12:18+00:00Codex-assisted visual audit valid_km {'x0': 74, 'x1': 697, 'y0': 11, 'y1': 336} X 0.0–96.0 / Y 0.0–1.0 P1 冻结原图审查;人工检查冻结原图刻度;ROI 四边按可见坐标轴框记录。

最近一次正式比较

Suite #3 · 2026-07-28T15:15:12+00:00

未通过

真值完整 117 / 120;通过 0;缺真值 0;无效案例 3。