Diffuse bone marrow edema is present in different pathological conditions including osteoporosis, fractures, osteonecrosis, and tumors. The MR imaging appearance is nonspecific in the absence of additional bone changes such as line fractures or cortical defects. Transient bone marrow edema is a syndrome that refers to any patient with reversible bone marrow edema on MR images [7]. The most frequently affected bones are the femoral head and neck and the distal femur where the bone marrow edema involves commonly the medial condyle. Usually, the patients do not remember a traumatic event. The etiology is unclear, but transient osteoporosis, neuromuscular dysfunctions, and transient ischemia are the probable causes for the presence of pain and edema [7]. However, some authors prefer the term of transient bone marrow edema to be used only in patients in whom osteopenia is not demonstrated on radiographs [7]. Clinically, the patients complain of pain, and the MR imaging shows a diffuse high-signal-intensity edema on T2-weighted images without any additional bone pathological changes (Fig. 11.5). A small amount of synovial fluid may be present. The diagnosis is based on the clinical findings, the absence of the previous trauma, the evolution of the pain, and the observation in MR imaging that the edema disappears spontaneously after a while.

Disuse osteopenia is a diffuse bone marrow change that appears after immobilization probably due to bone resorbtion and compromised architecture [8–10]. The changes are best seen on T2-weighted fat-suppressed images and may have different patterns from a generalized and diffuse high signal intensity relative to the hyaline cartilage lesions to a more “spotty” distribution of the lesions (Fig. 11.6) [8, 11]. The peak of the signal intensity changes is between 10 and 25 weeks after injury with the abnormalities disappearing after 65 weeks [11]. To recognize these changes is clinically important since the original bone strength is decreased during the presence of the bone marrow changes and there is an increased risk of fractures [12]. Although, on radiography, the disuse changes may mimic more aggressive pathologies as bone necrosis, multiple myeloma, or other malignant lesions, on MR imaging, the diagnosis is straightforward especially when correlated with the patient’s history.

The epiphyseal growth arrest lines represent low-signal-intensity linear changes of the subchondral bone and long bone metaphyses [13–15] (Fig. 11.7). They may be caused by immobilization during childhood, infection, or malnutrition [13–15]. MR imaging enables a better visualization of this bone-within-bone appearance compared to radiography [15].

The necrosis that involves the subchondral bone usually is referred to as avascular osteonecrosis, whereas necrosis of metaphyseal regions is referred to as bone infarction [16]. Spontaneous avascular osteonecrosis of the knee (SONK) is the second most common localization for osteonecrosis after the hip and affects mainly the medial condyle of elderly women over 55 years (Fig. 11.8) [17, 18]. Necrosis may be primary or secondary due to a variety of factors such as trauma (e.g., osteochondral fractures), systemic steroid therapy, alcoholism, or after arthroscopy. The secondary form of necrosis affects younger patients and may be bilateral [17]. In the early stage of osteonecrosis, MR imaging may show only presence of diffuse bone marrow edema within the medial condyle. In later stages, MR imaging demonstrates a serpiginous low-signal-intensity line in the medial condyle. The subchondral fragment may have similar signal intensity to the bone marrow or may display a hypointense signal on both T2-weighted and T1-weighted images with bone depression (Fig. 11.8).

Bone marrow infarction is usually asymptomatic and occurs predominantly in fatty marrow within metaphysis and diaphysis because of poor blood supply in comparison with the hematopoietic marrow which has a higher blood supply. The risk factors for bone infarction are sickle-cell disease, Gaucher’s disease, Hodgkin lymphoma, HIV infection, renal transplantation, myeloproliferative disorders, and chemotherapeutic and corticosteroid therapy (Figs. 11.3 and 11.9) [16, 18–21]. On T2-weighted images, the lesion is heterogeneously hyperintense and usually surrounded by an outer rim of low signal intensity with an inner rim of high signal intensity (“the double-line sign”) (Fig. 11.10) [16]. The outer rim is thought to represent sclerotic bone and the inner rim vascularized granulation tissue [16]. Very rarely, cyst formation in infarcted bones occurs [22].

The bone bruises or bone contusions are the result of direct or indirect trauma to the knee and are seen on MR images as subchondral, focal, ill-defined lesions of high signal intensity on T2-weighted images and low signal intensity on T1-weighted images. Usually, the adjacent bone contour is unaffected or may be discretely deepened. Pathologically, in the areas of contusions, edema, trabecular microfractures, and hemorrhage are present. In most of the cases, the bone bruises are secondary signs that accompany more severe lesions of the knee structures. Bone contusions may be single or multiple and are detected in 14–80 % of the patients with knee injuries [23]. The lateral tibial condyle is the most frequently involved area that is diagnosed on MR imaging (Fig. 11.11). When the bone contusions occur on both surfaces of the femoral and tibial articular surfaces, they are known as kissing contusions [24]. There are different injury patterns, and knowing the distribution of the lesions brings important information regarding the mechanism of the knee injury (Table 11.2). Moreover, the location of the bone bruise is an important information and may focus the attention to injuries that could be expected based on specific contusion patterns (Table 11.2). The strongest correlation between bone and ligamentous injuries is the correlation between the anterior cruciate ligament tear and the posterolateral tibial contusion alone or together with lateral femoral contusion (Fig. 11.12) [25].